diff --git a/README.md b/README.md
index 63b4f05..1f3d47d 100644
--- a/README.md
+++ b/README.md
@@ -8,25 +8,16 @@
   </picture>
 </p>
 
-
 <h3 align="center">
 Reliable and Efficient Semantic Prompt Caching
 </h3>
 <br>
 
-
-
-
 Semantic caching reduces LLM latency and cost by returning cached model responses for semantically similar prompts (not just exact matches). **vCache** is the first verified semantic cache that **guarantees user-defined error rate bounds**. vCache replaces static thresholds with **online-learned, embedding-specific decision boundaries**—no manual fine-tuning required. This enables reliable cached response reuse across any embedding model or workload.
 
-
-
 > [NOTE]
 > vCache is currently in active development. Features and APIs may change as we continue to improve the system.
 
-
-
-
 ## 🚀 Quick Install
 
 Install vCache in editable mode:
@@ -40,6 +31,7 @@ Then, set your OpenAI key:
 ```bash
 export OPENAI_API_KEY="your_api_key_here"
 ```
+
 (Note: vCache uses OpenAI by default for both LLM inference and embedding generation, but you can configure any other backend)
 
 Finally, use vCache in your Python code:
@@ -53,16 +45,14 @@ print(f"Response: {response}")
 ```
 
 By default, vCache uses:
+
 - `OpenAIInferenceEngine`
 - `OpenAIEmbeddingEngine`
 - `HNSWLibVectorDB`
-- `InMemoryEmbeddingMetadataStorage`
 - `NoEvictionPolicy`
 - `StringComparisonSimilarityEvaluator`
 - `VerifiedDecisionPolicy` with a maximum failure rate of 2%
 
-
-
 ## ⚙️ Advanced Configuration
 
 vCache is modular and highly configurable. Below is an example showing how to customize key components:
@@ -75,12 +65,12 @@ from vcache.main import VCache
 from vcache.config import VCacheConfig
 from vcache.inference_engine.strategies.open_ai import OpenAIInferenceEngine
 from vcache.vcache_core.cache.embedding_engine.strategies.open_ai import OpenAIEmbeddingEngine
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.strategies.in_memory import InMemoryEmbeddingMetadataStorage
 from vcache.vcache_core.similarity_evaluator.strategies.string_comparison import StringComparisonSimilarityEvaluator
 from vcache.vcache_policy.strategies.dynamic_local_threshold import VerifiedDecisionPolicy
 from vcache.vcache_policy.vcache_policy import VCachePolicy
-from vcache.vcache_core.cache.embedding_store.vector_db import HNSWLibVectorDB, SimilarityMetricType
+from vcache.vcache_core.cache.vector_db import HNSWLibVectorDB, SimilarityMetricType
 ```
+
 </details>
 
 ```python
@@ -92,7 +82,6 @@ vcache_config: VCacheConfig = VCacheConfig(
         similarity_metric_type=SimilarityMetricType.COSINE,
         max_capacity=100_000,
     ),
-    embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
     similarity_evaluator=StringComparisonSimilarityEvaluator,
 )
 
@@ -117,16 +106,16 @@ Semantic caching reduces LLM latency and cost by returning cached model response
 ### Architecture Overview
 
 1. **Embed & Store**  
-Each prompt is converted to a fixed-length vector (an “embedding”) and stored in a vector database along with its LLM response.
+   Each prompt is converted to a fixed-length vector (an “embedding”) and stored in a vector database along with its LLM response.
 
 2. **Nearest-Neighbor Lookup**  
-When a new prompt arrives, the cache embeds it and finds its most similar stored prompt using a similarity metric (e.g., cosine similarity).
+   When a new prompt arrives, the cache embeds it and finds its most similar stored prompt using a similarity metric (e.g., cosine similarity).
 
 3. **Similarity Score**  
-The system computes a score between 0 and 1 that quantifies how “close” the new prompt is to the retrieved entry.
+   The system computes a score between 0 and 1 that quantifies how “close” the new prompt is to the retrieved entry.
 
-4. **Decision: Exploit vs. Explore**  
-   - **Exploit (cache hit):** If the similarity is above a confidence bound, return the cached response.  
+4. **Decision: Exploit vs. Explore**
+   - **Exploit (cache hit):** If the similarity is above a confidence bound, return the cached response.
    - **Explore (cache miss):** Otherwise, infer the LLM for a response, add its embedding and answer to the cache, and return it.
 
 <p align="left">
@@ -139,6 +128,7 @@ The system computes a score between 0 and 1 that quantifies how “close” the
 </p>
 
 ### Why Fixed Thresholds Fall Short
+
 Existing semantic caches rely on a **global static threshold** to decide whether to reuse a cached response (exploit) or invoke the LLM (explore). If the similarity score exceeds this threshold, the cache reuses the response; otherwise, it infers the model. This strategy is fundamentally limited.
 
 - **Uniform threshold, diverse prompts:** A fixed threshold assumes all embeddings are equally distributed—ignoring that similarity is context-dependent.
@@ -161,11 +151,11 @@ vCache overcomes these limitations with two ideas:
 ### Benefits
 
 - **Reliability**  
-  Formally bounds the rate of incorrect cache hits to your chosen tolerance.  
+  Formally bounds the rate of incorrect cache hits to your chosen tolerance.
 - **Performance**  
-  Matches or exceeds static-threshold systems in cache hit rate and end-to-end latency.  
+  Matches or exceeds static-threshold systems in cache hit rate and end-to-end latency.
 - **Simplicity**  
-  Plug in any embedding model; vCache learns and adapts automatically at runtime.
+ Plug in any embedding model; vCache learns and adapts automatically at runtime.
 
   <p align="left">
   <picture>
@@ -178,25 +168,24 @@ vCache overcomes these limitations with two ideas:
 
 Please refer to the [vCache paper](https://arxiv.org/abs/2502.03771) for further details.
 
-
 ## 🛠 Developer Guide
 
 For advanced usage and development setup, see the [Developer Guide](ReadMe_Dev.md).
 
-
-
 ## 📊 Benchmarking vCache
 
 vCache includes a benchmarking framework to evaluate:
+
 - **Cache hit rate**
 - **Error rate**
 - **Latency improvement**
 - **...**
 
 We provide three open benchmarks:
-- **SemCacheLmArena** (chat-style prompts) - [Dataset  ↗](https://huggingface.co/datasets/vCache/SemBenchmarkLmArena)
-- **SemCacheClassification** (classification queries) - [Dataset  ↗](https://huggingface.co/datasets/vCache/SemBenchmarkClassification)
-- **SemCacheSearchQueries** (real-world search logs) - [Dataset  ↗](https://huggingface.co/datasets/vCache/SemBenchmarkSearchQueries)
+
+- **SemCacheLmArena** (chat-style prompts) - [Dataset ↗](https://huggingface.co/datasets/vCache/SemBenchmarkLmArena)
+- **SemCacheClassification** (classification queries) - [Dataset ↗](https://huggingface.co/datasets/vCache/SemBenchmarkClassification)
+- **SemCacheSearchQueries** (real-world search logs) - [Dataset ↗](https://huggingface.co/datasets/vCache/SemBenchmarkSearchQueries)
 
 See the [Benchmarking Documentation](benchmarks/ReadMe.md) for instructions.
 
@@ -211,4 +200,4 @@ If you use vCache for your research, please cite our [paper](https://arxiv.org/a
   journal={arXiv preprint arXiv:2502.03771},
   year={2025}
 }
-```
\ No newline at end of file
+```
diff --git a/benchmarks/benchmark.py b/benchmarks/benchmark.py
index 34efa30..0095c82 100644
--- a/benchmarks/benchmark.py
+++ b/benchmarks/benchmark.py
@@ -23,16 +23,13 @@
 from vcache.vcache_core.cache.embedding_engine.strategies.benchmark import (
     BenchmarkEmbeddingEngine,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage import (
-    InMemoryEmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db import (
+from vcache.vcache_core.cache.vector_db import (
     HNSWLibVectorDB,
     SimilarityMetricType,
 )
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
+    EmbeddingMetadataObj,
+)
 from vcache.vcache_core.similarity_evaluator import SimilarityEvaluator
 from vcache.vcache_core.similarity_evaluator.strategies.llm_comparison import (
     LLMComparisonSimilarityEvaluator,
@@ -396,7 +393,7 @@ def dump_results_to_json(self):
         var_ts_dict = {}
 
         metadata_objects: List[EmbeddingMetadataObj] = (
-            self.vcache.vcache_config.embedding_metadata_storage.get_all_embedding_metadata_objects()
+            self.vcache.vcache_config.vector_db.get_all_embedding_metadata_objects()
         )
 
         for metadata_object in metadata_objects:
@@ -486,7 +483,6 @@ def __run_baseline(
             similarity_metric_type=SimilarityMetricType.COSINE,
             max_capacity=MAX_VECTOR_DB_CAPACITY,
         ),
-        embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
         similarity_evaluator=similarity_evaluator,
     )
     vcache: VCache = VCache(vcache_config, vcache_policy)
diff --git a/test.py b/test.py
index 62b2da7..bf042b6 100644
--- a/test.py
+++ b/test.py
@@ -4,10 +4,7 @@
 from vcache.vcache_core.cache.embedding_engine.strategies.open_ai import (
     OpenAIEmbeddingEngine,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.strategies.in_memory import (
-    InMemoryEmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db import (
+from vcache.vcache_core.cache.vector_db import (
     HNSWLibVectorDB,
     SimilarityMetricType,
 )
@@ -27,7 +24,6 @@
         similarity_metric_type=SimilarityMetricType.COSINE,
         max_capacity=100000,
     ),
-    embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
     similarity_evaluator=StringComparisonSimilarityEvaluator,
 )
 vcache: VCache = VCache(vcache_config, vcache_policy)
diff --git a/tests/integration/test_concurrency.py b/tests/integration/test_concurrency.py
index 4174303..6a1a714 100644
--- a/tests/integration/test_concurrency.py
+++ b/tests/integration/test_concurrency.py
@@ -8,7 +8,6 @@
 
 from vcache import (
     HNSWLibVectorDB,
-    InMemoryEmbeddingMetadataStorage,
     LangChainEmbeddingEngine,
     StringComparisonSimilarityEvaluator,
     VCache,
@@ -46,7 +45,6 @@ def answers_similar(a, b):
                 model_name="sentence-transformers/all-mpnet-base-v2"
             ),
             vector_db=HNSWLibVectorDB(),
-            embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
             similarity_evaluator=similarity_evaluator,
         )
 
@@ -93,7 +91,9 @@ def do_inference(prompt):
                     time.sleep(1.5)
                     executor.map(do_inference, concurrent_prompts_chunk_2)
 
-        all_metadata_objects = vcache.vcache_config.embedding_metadata_storage.get_all_embedding_metadata_objects()
+        all_metadata_objects = (
+            vcache.vcache_config.vector_db.get_all_embedding_metadata_objects()
+        )
         final_observation_count = len(all_metadata_objects)
 
         for i, metadata_object in enumerate(all_metadata_objects):
diff --git a/tests/integration/test_dynamic_threshold.py b/tests/integration/test_dynamic_threshold.py
index 4d4d919..b04d147 100644
--- a/tests/integration/test_dynamic_threshold.py
+++ b/tests/integration/test_dynamic_threshold.py
@@ -4,7 +4,6 @@
 
 from vcache import (
     HNSWLibVectorDB,
-    InMemoryEmbeddingMetadataStorage,
     LangChainEmbeddingEngine,
     OpenAIInferenceEngine,
     VCache,
@@ -25,7 +24,6 @@ def create_default_config_and_policy():
             model_name="sentence-transformers/all-mpnet-base-v2"
         ),
         vector_db=HNSWLibVectorDB(),
-        embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
         system_prompt="Please answer in a single word with the first letter capitalized. Example: London",
     )
     policy = VerifiedDecisionPolicy(delta=0.05)
diff --git a/tests/integration/test_static_threshold.py b/tests/integration/test_static_threshold.py
index d75300e..96bdd5e 100644
--- a/tests/integration/test_static_threshold.py
+++ b/tests/integration/test_static_threshold.py
@@ -5,7 +5,6 @@
 from vcache import (
     BenchmarkStaticDecisionPolicy,
     HNSWLibVectorDB,
-    InMemoryEmbeddingMetadataStorage,
     LangChainEmbeddingEngine,
     OpenAIInferenceEngine,
     VCache,
@@ -25,7 +24,6 @@ def create_default_config_and_policy():
             model_name="sentence-transformers/all-mpnet-base-v2"
         ),
         vector_db=HNSWLibVectorDB(),
-        embedding_metadata_storage=InMemoryEmbeddingMetadataStorage(),
     )
     policy = BenchmarkStaticDecisionPolicy(threshold=0.8)
     return config, policy
diff --git a/tests/unit/EmbeddingMetadataStrategy/test_embedding_metadata.py b/tests/unit/EmbeddingMetadataStrategy/test_embedding_metadata.py
deleted file mode 100644
index e266159..0000000
--- a/tests/unit/EmbeddingMetadataStrategy/test_embedding_metadata.py
+++ /dev/null
@@ -1,32 +0,0 @@
-import unittest
-
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage import (
-    InMemoryEmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-
-
-class TestEmbeddingMetadataStorageStrategy(unittest.TestCase):
-    def test_in_memory_strategy(self):
-        embedding_metadata_storage = InMemoryEmbeddingMetadataStorage()
-
-        initial_obj = EmbeddingMetadataObj(embedding_id=0, response="test")
-        embedding_id = embedding_metadata_storage.add_metadata(
-            embedding_id=0, metadata=initial_obj
-        )
-        assert embedding_id == 0
-        assert embedding_metadata_storage.get_metadata(embedding_id=0) == initial_obj
-
-        updated_obj = EmbeddingMetadataObj(embedding_id=0, response="test2")
-        embedding_metadata_storage.update_metadata(embedding_id=0, metadata=updated_obj)
-        assert embedding_metadata_storage.get_metadata(embedding_id=0) == updated_obj
-
-        embedding_metadata_storage.flush()
-        with self.assertRaises(ValueError):
-            embedding_metadata_storage.get_metadata(embedding_id=0)
-
-
-if __name__ == "__main__":
-    unittest.main()
diff --git a/tests/unit/VCachePolicyStrategy/test_vcache_policy.py b/tests/unit/VCachePolicyStrategy/test_vcache_policy.py
index 50bdcbc..01d24c1 100644
--- a/tests/unit/VCachePolicyStrategy/test_vcache_policy.py
+++ b/tests/unit/VCachePolicyStrategy/test_vcache_policy.py
@@ -3,7 +3,7 @@
 from unittest.mock import MagicMock, patch
 
 from vcache.config import VCacheConfig
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
+from vcache.vcache_core.cache.vector_db import (
     EmbeddingMetadataObj,
 )
 from vcache.vcache_policy.strategies.verified import (
@@ -48,11 +48,9 @@ def update_metadata(embedding_id, embedding_metadata):
         mock_config = MagicMock(spec=VCacheConfig)
         mock_config.inference_engine = self.mock_inference_engine
         mock_config.similarity_evaluator = self.mock_similarity_evaluator
-        # Add all required attributes for Cache creation
-        mock_config.embedding_engine = MagicMock()
-        mock_config.embedding_metadata_storage = MagicMock()
         mock_config.vector_db = MagicMock()
         mock_config.eviction_policy = MagicMock()
+        mock_config.embedding_engine = MagicMock()
 
         self.policy = VerifiedDecisionPolicy()
         self.policy.setup(mock_config)
diff --git a/tests/unit/VectorDBStrategy/test_vector_db.py b/tests/unit/VectorDBStrategy/test_vector_db.py
index 2ca1de9..d5b3a2c 100644
--- a/tests/unit/VectorDBStrategy/test_vector_db.py
+++ b/tests/unit/VectorDBStrategy/test_vector_db.py
@@ -2,9 +2,8 @@
 
 import pytest
 
-from vcache.vcache_core.cache.embedding_store.vector_db import (
-    ChromaVectorDB,
-    FAISSVectorDB,
+from vcache.vcache_core.cache.vector_db import (
+    EmbeddingMetadataObj,
     HNSWLibVectorDB,
     SimilarityMetricType,
 )
@@ -12,10 +11,6 @@
 VECTOR_DB_PARAMS = [
     (HNSWLibVectorDB, SimilarityMetricType.COSINE),
     (HNSWLibVectorDB, SimilarityMetricType.EUCLIDEAN),
-    (FAISSVectorDB, SimilarityMetricType.COSINE),
-    (FAISSVectorDB, SimilarityMetricType.EUCLIDEAN),
-    (ChromaVectorDB, SimilarityMetricType.COSINE),
-    (ChromaVectorDB, SimilarityMetricType.EUCLIDEAN),
 ]
 
 
@@ -32,15 +27,19 @@ def test_add_and_get_knn(self, vector_db_class, similarity_metric_type):
 
         # Test with a single embedding
         embedding = [0.1, 0.2, 0.3]
-        id1 = vector_db.add(embedding=embedding)
+        metadata1 = EmbeddingMetadataObj(response="test response 1")
+        id1 = vector_db.add(embedding=embedding, metadata=metadata1)
+
         knn = vector_db.get_knn(embedding=embedding, k=1)
         assert len(knn) == 1
         assert abs(knn[0][0] - 1.0) < 1e-5  # Should be a perfect match
         assert knn[0][1] == id1
 
         # Test with multiple embeddings
-        vector_db.add(embedding=[0.2, 0.3, 0.4])
-        vector_db.add(embedding=[0.3, 0.4, 0.5])
+        metadata2 = EmbeddingMetadataObj(response="test response 2")
+        metadata3 = EmbeddingMetadataObj(response="test response 3")
+        vector_db.add(embedding=[0.2, 0.3, 0.4], metadata=metadata2)
+        vector_db.add(embedding=[0.3, 0.4, 0.5], metadata=metadata3)
 
         # Verify we get all embeddings when k is large enough
         knn = vector_db.get_knn(embedding=embedding, k=3)
@@ -59,8 +58,10 @@ def test_remove(self, vector_db_class, similarity_metric_type):
         vector_db = vector_db_class(similarity_metric_type=similarity_metric_type)
 
         # Add multiple embeddings
-        id1 = vector_db.add(embedding=[0.1, 0.2, 0.3])
-        id2 = vector_db.add(embedding=[0.2, 0.3, 0.4])
+        metadata1 = EmbeddingMetadataObj(response="test response 1")
+        metadata2 = EmbeddingMetadataObj(response="test response 2")
+        id1 = vector_db.add(embedding=[0.1, 0.2, 0.3], metadata=metadata1)
+        id2 = vector_db.add(embedding=[0.2, 0.3, 0.4], metadata=metadata2)
 
         # Verify both exist
         knn = vector_db.get_knn(embedding=[0.1, 0.2, 0.3], k=2)
@@ -83,9 +84,12 @@ def test_reset(self, vector_db_class, similarity_metric_type):
         vector_db = vector_db_class(similarity_metric_type=similarity_metric_type)
 
         # Add multiple embeddings
-        vector_db.add(embedding=[0.1, 0.2, 0.3])
-        vector_db.add(embedding=[0.2, 0.3, 0.4])
-        vector_db.add(embedding=[0.3, 0.4, 0.5])
+        metadata1 = EmbeddingMetadataObj(response="test response 1")
+        metadata2 = EmbeddingMetadataObj(response="test response 2")
+        metadata3 = EmbeddingMetadataObj(response="test response 3")
+        vector_db.add(embedding=[0.1, 0.2, 0.3], metadata=metadata1)
+        vector_db.add(embedding=[0.2, 0.3, 0.4], metadata=metadata2)
+        vector_db.add(embedding=[0.3, 0.4, 0.5], metadata=metadata3)
 
         # Verify embeddings exist
         knn = vector_db.get_knn(embedding=[0.1, 0.2, 0.3], k=3)
@@ -98,6 +102,37 @@ def test_reset(self, vector_db_class, similarity_metric_type):
         knn = vector_db.get_knn(embedding=[0.1, 0.2, 0.3], k=3)
         assert len(knn) == 0
 
+    @pytest.mark.parametrize(
+        "vector_db_class, similarity_metric_type",
+        VECTOR_DB_PARAMS,
+    )
+    def test_metadata_operations(self, vector_db_class, similarity_metric_type):
+        """Test metadata operations of the vector database."""
+        vector_db = vector_db_class(similarity_metric_type=similarity_metric_type)
+
+        # Add embedding with metadata (embedding_id will be set automatically)
+        metadata = EmbeddingMetadataObj(response="test response")
+        embedding_id = vector_db.add(embedding=[0.1, 0.2, 0.3], metadata=metadata)
+
+        # Test get metadata
+        retrieved_metadata = vector_db.get_metadata(embedding_id)
+        assert retrieved_metadata.response == "test response"
+        assert (
+            retrieved_metadata.embedding_id == embedding_id
+        )  # Should be set automatically
+
+        # Test update metadata
+        updated_metadata = EmbeddingMetadataObj(response="updated response")
+        vector_db.update_metadata(embedding_id, updated_metadata)
+
+        retrieved_metadata = vector_db.get_metadata(embedding_id)
+        assert retrieved_metadata.response == "updated response"
+
+        # Test get all metadata objects
+        all_metadata = vector_db.get_all_embedding_metadata_objects()
+        assert len(all_metadata) == 1
+        assert all_metadata[0].response == "updated response"
+
 
 if __name__ == "__main__":
     unittest.main()
diff --git a/vcache/__init__.py b/vcache/__init__.py
index e5140a1..8e989b4 100644
--- a/vcache/__init__.py
+++ b/vcache/__init__.py
@@ -20,27 +20,19 @@
     OpenAIEmbeddingEngine,
 )
 
-# Embedding metadata storage
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-    InMemoryEmbeddingMetadataStorage,
+# Eviction policies
+from vcache.vcache_core.cache.eviction_policy import (
+    EvictionPolicy,
+    LRUEvictionPolicy,
 )
 
 # Vector databases
-from vcache.vcache_core.cache.embedding_store.vector_db import (
-    ChromaVectorDB,
-    FAISSVectorDB,
+from vcache.vcache_core.cache.vector_db import (
     HNSWLibVectorDB,
     SimilarityMetricType,
     VectorDB,
 )
 
-# Eviction policies
-from vcache.vcache_core.cache.eviction_policy import (
-    EvictionPolicy,
-    LRUEvictionPolicy,
-)
-
 # Similarity evaluators
 from vcache.vcache_core.similarity_evaluator import (
     SimilarityEvaluator,
@@ -71,9 +63,7 @@
     "LangChainEmbeddingEngine",
     # Vector databases
     "VectorDB",
-    "FAISSVectorDB",
     "HNSWLibVectorDB",
-    "ChromaVectorDB",
     "SimilarityMetricType",
     # Similarity evaluators
     "SimilarityEvaluator",
@@ -81,9 +71,6 @@
     # Eviction policies
     "EvictionPolicy",
     "LRUEvictionPolicy",
-    # Embedding metadata storage
-    "EmbeddingMetadataStorage",
-    "InMemoryEmbeddingMetadataStorage",
     # vCache Policies
     "VCachePolicy",
     "VerifiedDecisionPolicy",
diff --git a/vcache/config.py b/vcache/config.py
index f337648..d284920 100644
--- a/vcache/config.py
+++ b/vcache/config.py
@@ -4,20 +4,14 @@
 from vcache.inference_engine.strategies.open_ai import OpenAIInferenceEngine
 from vcache.vcache_core.cache.embedding_engine import OpenAIEmbeddingEngine
 from vcache.vcache_core.cache.embedding_engine.embedding_engine import EmbeddingEngine
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.strategies.in_memory import (
-    InMemoryEmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db import VectorDB
-from vcache.vcache_core.cache.embedding_store.vector_db.strategies.hnsw_lib import (
-    HNSWLibVectorDB,
-)
 from vcache.vcache_core.cache.eviction_policy.eviction_policy import EvictionPolicy
 from vcache.vcache_core.cache.eviction_policy.strategies.no_eviction import (
     NoEvictionPolicy,
 )
+from vcache.vcache_core.cache.vector_db.strategies.hnsw_lib import (
+    HNSWLibVectorDB,
+)
+from vcache.vcache_core.cache.vector_db.vector_db import VectorDB
 from vcache.vcache_core.similarity_evaluator.similarity_evaluator import (
     SimilarityEvaluator,
 )
@@ -36,7 +30,6 @@ def __init__(
         inference_engine: InferenceEngine = OpenAIInferenceEngine(),
         embedding_engine: EmbeddingEngine = OpenAIEmbeddingEngine(),
         vector_db: VectorDB = HNSWLibVectorDB(),
-        embedding_metadata_storage: EmbeddingMetadataStorage = InMemoryEmbeddingMetadataStorage(),
         eviction_policy: EvictionPolicy = NoEvictionPolicy(),
         similarity_evaluator: SimilarityEvaluator = StringComparisonSimilarityEvaluator(),
         system_prompt: Optional[str] = None,
@@ -47,8 +40,7 @@ def __init__(
         Args:
             inference_engine: Engine for generating responses from prompts.
             embedding_engine: Engine for generating embeddings from text.
-            vector_db: Vector database for storing and retrieving embeddings.
-            embedding_metadata_storage: Storage for embedding metadata.
+            vector_db: Vector database for storing embeddings and metadata.
             eviction_policy: Policy for removing items from cache when full.
             similarity_evaluator: Evaluator for determining similarity between prompts.
             system_prompt: Optional system prompt to use for all inferences.
@@ -57,7 +49,6 @@ def __init__(
         self.embedding_engine = embedding_engine
         self.vector_db = vector_db
         self.eviction_policy = eviction_policy
-        self.embedding_metadata_storage = embedding_metadata_storage
         self.similarity_evaluator = similarity_evaluator
         self.similarity_evaluator.set_inference_engine(self.inference_engine)
         self.system_prompt = system_prompt
diff --git a/vcache/vcache_core/cache/__init__.py b/vcache/vcache_core/cache/__init__.py
index 861d231..7882f73 100644
--- a/vcache/vcache_core/cache/__init__.py
+++ b/vcache/vcache_core/cache/__init__.py
@@ -1,21 +1,15 @@
 from vcache.vcache_core.cache.cache import Cache
 from vcache.vcache_core.cache.embedding_engine.embedding_engine import EmbeddingEngine
-from vcache.vcache_core.cache.embedding_store import EmbeddingStore
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
 from vcache.vcache_core.cache.eviction_policy.eviction_policy import EvictionPolicy
 from vcache.vcache_core.cache.eviction_policy.system_monitor import SystemMonitor
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
+    EmbeddingMetadataObj,
+)
 
 __all__ = [
     "Cache",
-    "EmbeddingStore",
     "EmbeddingMetadataObj",
     "EmbeddingEngine",
-    "EmbeddingMetadataStorage",
     "EvictionPolicy",
     "SystemMonitor",
 ]
diff --git a/vcache/vcache_core/cache/cache.py b/vcache/vcache_core/cache/cache.py
index 60adf17..7e29a4e 100644
--- a/vcache/vcache_core/cache/cache.py
+++ b/vcache/vcache_core/cache/cache.py
@@ -1,11 +1,11 @@
 from typing import List
 
 from vcache.vcache_core.cache.embedding_engine.embedding_engine import EmbeddingEngine
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
+from vcache.vcache_core.cache.eviction_policy.eviction_policy import EvictionPolicy
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
     EmbeddingMetadataObj,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
-from vcache.vcache_core.cache.eviction_policy.eviction_policy import EvictionPolicy
+from vcache.vcache_core.cache.vector_db.vector_db import VectorDB
 
 
 class Cache:
@@ -15,29 +15,25 @@ class Cache:
 
     def __init__(
         self,
-        embedding_store: EmbeddingStore,
+        vector_db: VectorDB,
         embedding_engine: EmbeddingEngine,
         eviction_policy: EvictionPolicy,
     ):
         """
-        Initialize cache with embedding store, engine, and eviction policy.
+        Initialize cache with vector database, engine, and eviction policy.
 
         Args:
-            embedding_store: Store for managing embeddings and metadata.
+            vector_db: Vector database for managing embeddings and metadata.
             embedding_engine: Engine for generating embeddings from text.
             eviction_policy: Policy for removing items when cache is full.
         """
-        self.embedding_store = embedding_store
+        self.vector_db = vector_db
         self.embedding_engine = embedding_engine
         self.eviction_policy = eviction_policy
 
     def add(self, prompt: str, response: str) -> int:
         """
-        Compute the embedding for the prompt, add an embedding to the vector database and a new metadata object.
-
-        IMPORTANT: The embedding is computed first and then added to the vector database.
-        The metadata object is added last.
-        Consider this when implementing asynchronous logic to prevent race conditions.
+        Compute the embedding for the prompt, add an embedding to the vector database with metadata.
 
         Args:
             prompt: The prompt to add to the cache.
@@ -47,7 +43,9 @@ def add(self, prompt: str, response: str) -> int:
             The id of the embedding.
         """
         embedding = self.embedding_engine.get_embedding(prompt)
-        return self.embedding_store.add_embedding(embedding, response)
+        metadata = EmbeddingMetadataObj(response=response)
+        embedding_id = self.vector_db.add(embedding, metadata)
+        return embedding_id
 
     def remove(self, embedding_id: int) -> int:
         """
@@ -59,7 +57,7 @@ def remove(self, embedding_id: int) -> int:
         Returns:
             The id of the embedding.
         """
-        return self.embedding_store.remove(embedding_id)
+        return self.vector_db.remove(embedding_id)
 
     def get_knn(self, prompt: str, k: int) -> List[tuple[float, int]]:
         """
@@ -73,13 +71,13 @@ def get_knn(self, prompt: str, k: int) -> List[tuple[float, int]]:
             A list of tuples, each containing a similarity score and an embedding id.
         """
         embedding = self.embedding_engine.get_embedding(prompt)
-        return self.embedding_store.get_knn(embedding, k)
+        return self.vector_db.get_knn(embedding, k)
 
     def flush(self) -> None:
         """
         Flush all data from the cache.
         """
-        self.embedding_store.reset()
+        self.vector_db.reset()
 
     def get_metadata(self, embedding_id: int) -> EmbeddingMetadataObj:
         """
@@ -91,7 +89,7 @@ def get_metadata(self, embedding_id: int) -> EmbeddingMetadataObj:
         Returns:
             The metadata of the embedding.
         """
-        return self.embedding_store.get_metadata(embedding_id)
+        return self.vector_db.get_metadata(embedding_id)
 
     def update_metadata(
         self, embedding_id: int, embedding_metadata: EmbeddingMetadataObj
@@ -106,7 +104,7 @@ def update_metadata(
         Returns:
             The updated metadata of the embedding.
         """
-        return self.embedding_store.update_metadata(embedding_id, embedding_metadata)
+        return self.vector_db.update_metadata(embedding_id, embedding_metadata)
 
     def get_current_capacity(self) -> int:
         """
@@ -125,7 +123,7 @@ def is_empty(self) -> bool:
         Returns:
             True if the cache is empty, False otherwise.
         """
-        return self.embedding_store.is_empty()
+        return self.vector_db.is_empty()
 
     def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
         """
@@ -134,4 +132,4 @@ def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
         Returns:
             A list of all the embedding metadata objects in the cache.
         """
-        return self.embedding_store.embedding_metadata_storage.get_all_embedding_metadata_objects()
+        return self.vector_db.get_all_embedding_metadata_objects()
diff --git a/vcache/vcache_core/cache/embedding_store/__init__.py b/vcache/vcache_core/cache/embedding_store/__init__.py
deleted file mode 100644
index febd930..0000000
--- a/vcache/vcache_core/cache/embedding_store/__init__.py
+++ /dev/null
@@ -1,11 +0,0 @@
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
-from vcache.vcache_core.cache.embedding_store.vector_db import (
-    SimilarityMetricType,
-    VectorDB,
-)
-
-__all__ = [
-    "VectorDB",
-    "EmbeddingStore",
-    "SimilarityMetricType",
-]
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/__init__.py b/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/__init__.py
deleted file mode 100644
index 77692de..0000000
--- a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/__init__.py
+++ /dev/null
@@ -1,8 +0,0 @@
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.strategies.in_memory import (
-    InMemoryEmbeddingMetadataStorage,
-)
-
-__all__ = ["EmbeddingMetadataStorage", "InMemoryEmbeddingMetadataStorage"]
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_storage.py b/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_storage.py
deleted file mode 100644
index 0c7fcbd..0000000
--- a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_storage.py
+++ /dev/null
@@ -1,82 +0,0 @@
-from abc import ABC, abstractmethod
-from typing import List
-
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-
-
-class EmbeddingMetadataStorage(ABC):
-    """
-    Abstract base class for embedding metadata storage.
-    """
-
-    @abstractmethod
-    def add_metadata(self, embedding_id: int, metadata: EmbeddingMetadataObj) -> int:
-        """
-        Add metadata for a specific embedding.
-
-        Args:
-            embedding_id: The id of the embedding to add the metadata for.
-            metadata: The metadata to add to the embedding.
-
-        Returns:
-            The id of the embedding.
-        """
-        pass
-
-    @abstractmethod
-    def get_metadata(self, embedding_id: int) -> EmbeddingMetadataObj:
-        """
-        Get metadata for a specific embedding.
-
-        Args:
-            embedding_id: The id of the embedding to get the metadata for.
-
-        Returns:
-            The metadata of the embedding.
-        """
-        pass
-
-    @abstractmethod
-    def update_metadata(
-        self, embedding_id: int, metadata: EmbeddingMetadataObj
-    ) -> EmbeddingMetadataObj:
-        """
-        Update metadata for a specific embedding.
-
-        Args:
-            embedding_id: The id of the embedding to update the metadata for.
-            metadata: The metadata to update the embedding with.
-
-        Returns:
-            The updated metadata of the embedding.
-        """
-        pass
-
-    @abstractmethod
-    def remove_metadata(self, embedding_id: int) -> None:
-        """
-        Remove metadata for a specific embedding.
-
-        Args:
-            embedding_id: The id of the embedding to remove the metadata for.
-        """
-        pass
-
-    @abstractmethod
-    def flush(self) -> None:
-        """
-        Flush all metadata from storage.
-        """
-        pass
-
-    @abstractmethod
-    def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
-        """
-        Get all embedding metadata objects in storage.
-
-        Returns:
-            A list of all the embedding metadata objects in the storage.
-        """
-        pass
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/__init__.py b/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/__init__.py
deleted file mode 100644
index e69de29..0000000
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/in_memory.py b/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/in_memory.py
deleted file mode 100644
index 3e49c8e..0000000
--- a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/in_memory.py
+++ /dev/null
@@ -1,110 +0,0 @@
-from typing import Any, Dict, List, Optional
-
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
-
-
-class InMemoryEmbeddingMetadataStorage(EmbeddingMetadataStorage):
-    """
-    In-memory implementation of embedding metadata storage.
-    """
-
-    def __init__(self):
-        """
-        Initialize in-memory embedding metadata storage.
-        """
-        self.metadata_storage: Dict[int, "EmbeddingMetadataObj"] = {}
-
-    def add_metadata(
-        self, embedding_id: int, metadata: Optional[Dict[str, Any]] = None
-    ) -> None:
-        """
-        Add metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding to add metadata for.
-            metadata: The metadata to add.
-
-        Returns:
-            The embedding ID.
-        """
-        self.metadata_storage[embedding_id] = metadata
-        return embedding_id
-
-    def get_metadata(self, embedding_id: int) -> Optional[Dict[str, Any]]:
-        """
-        Get metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding to get metadata for.
-
-        Returns:
-            The metadata for the embedding.
-
-        Raises:
-            ValueError: If embedding metadata is not found.
-        """
-        if embedding_id not in self.metadata_storage:
-            raise ValueError(
-                f"Embedding metadata for embedding id {embedding_id} not found"
-            )
-        else:
-            return self.metadata_storage[embedding_id]
-
-    def update_metadata(
-        self, embedding_id: int, metadata: Optional[Dict[str, Any]] = None
-    ) -> bool:
-        """
-        Update metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding to update metadata for.
-            metadata: The new metadata to set.
-
-        Returns:
-            The updated metadata.
-
-        Raises:
-            ValueError: If embedding metadata is not found.
-        """
-        if embedding_id not in self.metadata_storage:
-            raise ValueError(
-                f"Embedding metadata for embedding id {embedding_id} not found"
-            )
-        else:
-            self.metadata_storage[embedding_id] = metadata
-            return metadata
-
-    def remove_metadata(self, embedding_id: int) -> bool:
-        """
-        Remove metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding to remove metadata for.
-
-        Returns:
-            True if metadata was removed, False if not found.
-        """
-        if embedding_id in self.metadata_storage:
-            del self.metadata_storage[embedding_id]
-            return True
-        return False
-
-    def flush(self) -> None:
-        """
-        Flush all metadata from storage.
-        """
-        self.metadata_storage = {}
-
-    def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
-        """
-        Get all embedding metadata objects in storage.
-
-        Returns:
-            A list of all embedding metadata objects.
-        """
-        return list(self.metadata_storage.values())
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/langchain.py b/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/langchain.py
deleted file mode 100644
index 10025fb..0000000
--- a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/strategies/langchain.py
+++ /dev/null
@@ -1,93 +0,0 @@
-from typing import Any, Dict, List, Optional
-
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
-
-
-class LangchainMetadataStorage(EmbeddingMetadataStorage):
-    """
-    LangChain-based metadata storage implementation (placeholder).
-    """
-
-    def __init__(self):
-        """
-        Initialize LangChain metadata storage.
-        """
-        # TODO
-        pass
-
-    def add_metadata(
-        self, embedding_id: int, metadata: Optional[Dict[str, Any]] = None
-    ) -> None:
-        """
-        Add metadata for an embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-            metadata: The metadata to add.
-        """
-        # TODO
-        pass
-
-    def get_metadata(self, embedding_id: int) -> Optional[Dict[str, Any]]:
-        """
-        Get metadata for an embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-
-        Returns:
-            The metadata for the embedding, or None if not found.
-        """
-        # TODO
-        pass
-
-    def update_metadata(
-        self, embedding_id: int, metadata: Optional[Dict[str, Any]] = None
-    ) -> bool:
-        """
-        Update metadata for an embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-            metadata: The new metadata.
-
-        Returns:
-            True if the update was successful, False otherwise.
-        """
-        # TODO
-        pass
-
-    def remove_metadata(self, embedding_id: int) -> bool:
-        """
-        Remove metadata for an embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-
-        Returns:
-            True if the removal was successful, False otherwise.
-        """
-        # TODO
-        pass
-
-    def flush(self) -> None:
-        """
-        Flush any pending changes to storage.
-        """
-        # TODO
-        pass
-
-    def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
-        """
-        Get all embedding metadata objects.
-
-        Returns:
-            List of all embedding metadata objects.
-        """
-        # TODO
-        pass
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_store.py b/vcache/vcache_core/cache/embedding_store/embedding_store.py
deleted file mode 100644
index c36d10e..0000000
--- a/vcache/vcache_core/cache/embedding_store/embedding_store.py
+++ /dev/null
@@ -1,139 +0,0 @@
-import threading
-from typing import List
-
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage import (
-    EmbeddingMetadataStorage,
-)
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
-    EmbeddingMetadataObj,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db.vector_db import VectorDB
-
-
-class EmbeddingStore:
-    """
-    Store for managing embeddings and their associated metadata.
-    """
-
-    def __init__(
-        self,
-        vector_db: VectorDB,
-        embedding_metadata_storage: EmbeddingMetadataStorage,
-    ):
-        """
-        Initialize embedding store with vector database and metadata storage.
-
-        Args:
-            vector_db: Vector database for storing embeddings.
-            embedding_metadata_storage: Storage for embedding metadata.
-        """
-        self.vector_db = vector_db
-        self.embedding_metadata_storage = embedding_metadata_storage
-        self._add_lock = threading.Lock()
-        self._remove_lock = threading.Lock()
-
-    def add_embedding(self, embedding: List[float], response: str) -> int:
-        """
-        Add an embedding to the vector database and a new metadata object.
-
-        This operation is thread-safe.
-
-        Args:
-            embedding: The embedding vector to add.
-            response: The response associated with the embedding.
-
-        Returns:
-            The ID of the added embedding.
-        """
-        with self._add_lock:
-            embedding_id = self.vector_db.add(embedding)
-            metadata = EmbeddingMetadataObj(
-                embedding_id=embedding_id,
-                response=response,
-            )
-            self.embedding_metadata_storage.add_metadata(
-                embedding_id=embedding_id, metadata=metadata
-            )
-            return embedding_id
-
-    def remove(self, embedding_id: int) -> int:
-        """
-        Remove an embedding and its metadata from the store.
-
-        This operation is thread-safe.
-
-        Args:
-            embedding_id: The ID of the embedding to remove.
-
-        Returns:
-            The ID of the removed embedding.
-        """
-        with self._remove_lock:
-            self.embedding_metadata_storage.remove_metadata(embedding_id)
-            return self.vector_db.remove(embedding_id)
-
-    def get_knn(self, embedding: List[float], k: int) -> List[tuple[float, int]]:
-        """
-        Get k-nearest neighbors for the given embedding.
-
-        Args:
-            embedding: The embedding to find neighbors for.
-            k: The number of neighbors to return.
-
-        Returns:
-            List of tuples containing similarity scores and embedding IDs.
-        """
-        return self.vector_db.get_knn(embedding, k)
-
-    def reset(self) -> None:
-        """
-        Reset the embedding store to empty state.
-        """
-        self.embedding_metadata_storage.flush()
-        return self.vector_db.reset()
-
-    def calculate_storage_consumption(self) -> int:
-        """
-        Calculate the storage consumption of the embedding store.
-
-        Returns:
-            The storage consumption in bytes.
-        """
-        # TODO: Add metadata logic
-        return -1
-
-    def get_metadata(self, embedding_id: int) -> "EmbeddingMetadataObj":
-        """
-        Get metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-
-        Returns:
-            The metadata object for the embedding.
-        """
-        return self.embedding_metadata_storage.get_metadata(embedding_id)
-
-    def update_metadata(
-        self, embedding_id: int, metadata: "EmbeddingMetadataObj"
-    ) -> "EmbeddingMetadataObj":
-        """
-        Update metadata for a specific embedding.
-
-        Args:
-            embedding_id: The ID of the embedding.
-            metadata: The new metadata object.
-
-        Returns:
-            The updated metadata object.
-        """
-        return self.embedding_metadata_storage.update_metadata(embedding_id, metadata)
-
-    def is_empty(self) -> bool:
-        """
-        Check if the embedding store is empty.
-
-        Returns:
-            True if the store is empty, False otherwise.
-        """
-        return self.vector_db.is_empty()
diff --git a/vcache/vcache_core/cache/embedding_store/vector_db/__init__.py b/vcache/vcache_core/cache/embedding_store/vector_db/__init__.py
deleted file mode 100644
index bb4a6c3..0000000
--- a/vcache/vcache_core/cache/embedding_store/vector_db/__init__.py
+++ /dev/null
@@ -1,21 +0,0 @@
-from vcache.vcache_core.cache.embedding_store.vector_db.strategies.chroma import (
-    ChromaVectorDB,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db.strategies.faiss import (
-    FAISSVectorDB,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db.strategies.hnsw_lib import (
-    HNSWLibVectorDB,
-)
-from vcache.vcache_core.cache.embedding_store.vector_db.vector_db import (
-    SimilarityMetricType,
-    VectorDB,
-)
-
-__all__ = [
-    "VectorDB",
-    "SimilarityMetricType",
-    "HNSWLibVectorDB",
-    "FAISSVectorDB",
-    "ChromaVectorDB",
-]
diff --git a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/chroma.py b/vcache/vcache_core/cache/embedding_store/vector_db/strategies/chroma.py
deleted file mode 100644
index f2cfe47..0000000
--- a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/chroma.py
+++ /dev/null
@@ -1,140 +0,0 @@
-from typing import List
-
-import chromadb
-
-from vcache.vcache_core.cache.embedding_store.vector_db.vector_db import (
-    SimilarityMetricType,
-    VectorDB,
-)
-
-
-class ChromaVectorDB(VectorDB):
-    """
-    ChromaDB-based vector database implementation for efficient similarity search.
-    """
-
-    def __init__(
-        self, similarity_metric_type: SimilarityMetricType = SimilarityMetricType.COSINE
-    ):
-        """
-        Initialize ChromaDB vector database.
-
-        Args:
-            similarity_metric_type: The similarity metric to use for comparisons.
-        """
-        self.__next_embedding_id = 0
-        self.collection = None
-        self.client = None
-        self.similarity_metric_type = similarity_metric_type
-
-    def add(self, embedding: List[float]) -> int:
-        """
-        Add an embedding vector to the database.
-
-        Args:
-            embedding: The embedding vector to add.
-
-        Returns:
-            The unique ID assigned to the added embedding.
-        """
-        if self.collection is None:
-            self._init_vector_store(len(embedding))
-        id = self.__next_embedding_id
-        self.collection.add(embeddings=[embedding], ids=[str(id)])
-        self.__next_embedding_id += 1
-        return id
-
-    def remove(self, embedding_id: int) -> int:
-        """
-        Remove an embedding from the database.
-
-        Args:
-            embedding_id: The ID of the embedding to remove.
-
-        Returns:
-            The ID of the removed embedding.
-
-        Raises:
-            ValueError: If the collection is not initialized.
-        """
-        if self.collection is None:
-            raise ValueError("Collection is not initialized")
-        self.collection.delete(ids=[str(embedding_id)])
-        return embedding_id
-
-    def get_knn(self, embedding: List[float], k: int) -> List[tuple[float, int]]:
-        """
-        Get k-nearest neighbors for the given embedding.
-
-        Args:
-            embedding: The query embedding vector.
-            k: The number of nearest neighbors to return.
-
-        Returns:
-            List of tuples containing similarity scores and embedding IDs.
-
-        Raises:
-            ValueError: If the collection is not initialized.
-        """
-        if self.collection is None:
-            raise ValueError("Collection is not initialized")
-        if self.collection.count() == 0:
-            return []
-        k_ = min(k, self.collection.count())
-        results = self.collection.query(
-            query_embeddings=[embedding], n_results=k_, include=["distances"]
-        )
-        distances = results.get("distances", [[]])[0]
-        ids = results.get("ids", [[]])[0]
-        return [
-            (
-                self.transform_similarity_score(
-                    float(dist), self.similarity_metric_type.value
-                ),
-                int(idx),
-            )
-            for dist, idx in zip(distances, ids)
-        ]
-
-    def reset(self) -> None:
-        """
-        Reset the vector database to empty state.
-        """
-        if self.collection is not None:
-            self.collection.delete(ids=self.collection.get()["ids"])
-        self.__next_embedding_id = 0
-
-    def _init_vector_store(self, embedding_dim: int):
-        """
-        Initialize the ChromaDB collection with the given embedding dimension.
-
-        Args:
-            embedding_dim: The dimension of the embedding vectors.
-
-        Raises:
-            ValueError: If the similarity metric type is invalid.
-        """
-        self.client = chromadb.Client()
-        collection_name = f"vcache_collection_{id(self)}"
-        metric_type = self.similarity_metric_type.value
-        match metric_type:
-            case "cosine":
-                space = "cosine"
-            case "euclidean":
-                space = "l2"
-            case _:
-                raise ValueError(f"Invalid similarity metric type: {metric_type}")
-        self.collection = self.client.create_collection(
-            name=collection_name,
-            metadata={"dimension": embedding_dim, "hnsw:space": space},
-            get_or_create=True,
-        )
-
-    def is_empty(self) -> bool:
-        """
-        Check if the vector database is empty.
-
-        Returns:
-            True if the database contains no embeddings, False otherwise.
-        """
-        return self.collection.count() == 0
diff --git a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/faiss.py b/vcache/vcache_core/cache/embedding_store/vector_db/strategies/faiss.py
deleted file mode 100644
index ad6f883..0000000
--- a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/faiss.py
+++ /dev/null
@@ -1,170 +0,0 @@
-from typing import List
-
-import faiss
-import numpy as np
-
-from vcache.vcache_core.cache.embedding_store.vector_db.vector_db import (
-    SimilarityMetricType,
-    VectorDB,
-)
-
-
-class FAISSVectorDB(VectorDB):
-    """
-    FAISS-based vector database implementation for efficient similarity search.
-    """
-
-    def __init__(
-        self, similarity_metric_type: SimilarityMetricType = SimilarityMetricType.COSINE
-    ):
-        """
-        Initialize FAISS vector database.
-
-        Args:
-            similarity_metric_type: The similarity metric to use for comparisons.
-        """
-        self.similarity_metric_type = similarity_metric_type
-        self.__next_embedding_id = 0
-        self.index = None
-
-    def transform_similarity_score(
-        self, similarity_score: float, metric_type: str
-    ) -> float:
-        """
-        Transform similarity score based on the metric type.
-
-        Args:
-            similarity_score: The raw similarity score.
-            metric_type: The type of similarity metric used.
-
-        Returns:
-            The transformed similarity score.
-
-        Raises:
-            ValueError: If the similarity metric type is invalid.
-        """
-        match metric_type:
-            case "cosine":
-                return similarity_score
-            case "euclidean":
-                return 1 - similarity_score
-            case _:
-                raise ValueError(f"Invalid similarity metric type: {metric_type}")
-
-    def add(self, embedding: List[float]) -> int:
-        """
-        Add an embedding vector to the database.
-
-        Args:
-            embedding: The embedding vector to add.
-
-        Returns:
-            The unique ID assigned to the added embedding.
-        """
-        if self.index is None:
-            self._init_vector_store(len(embedding))
-        id = self.__next_embedding_id
-        ids = np.array([id], dtype=np.int64)
-        embedding_array = np.array([embedding], dtype=np.float32)
-        metric_type = self.similarity_metric_type.value
-        # Normalize the embedding vector if the metric type is cosine
-        if metric_type == "cosine":
-            faiss.normalize_L2(embedding_array)
-        self.index.add_with_ids(embedding_array, ids)
-        self.__next_embedding_id += 1
-        return id
-
-    def remove(self, embedding_id: int) -> int:
-        """
-        Remove an embedding from the database.
-
-        Args:
-            embedding_id: The ID of the embedding to remove.
-
-        Returns:
-            The ID of the removed embedding.
-
-        Raises:
-            ValueError: If the index is not initialized.
-        """
-        if self.index is None:
-            raise ValueError("Index is not initialized")
-        id_array = np.array([embedding_id], dtype=np.int64)
-        self.index.remove_ids(
-            faiss.IDSelectorBatch(id_array.size, faiss.swig_ptr(id_array))
-        )
-        return embedding_id
-
-    def get_knn(self, embedding: List[float], k: int) -> List[tuple[float, int]]:
-        """
-        Get k-nearest neighbors for the given embedding.
-
-        Args:
-            embedding: The query embedding vector.
-            k: The number of nearest neighbors to return.
-
-        Returns:
-            List of tuples containing similarity scores and embedding IDs.
-
-        Raises:
-            ValueError: If the index is not initialized.
-        """
-        if self.index is None:
-            raise ValueError("Index is not initialized")
-        if self.index.ntotal == 0:
-            return []
-        k_ = min(k, self.index.ntotal)
-        query_vector = np.array([embedding], dtype=np.float32)
-        metric_type = self.similarity_metric_type.value
-        # Normalize the query vector if the metric type is cosine
-        if metric_type == "cosine":
-            faiss.normalize_L2(query_vector)
-        distances, indices = self.index.search(query_vector, k_)
-        # Filter out results where index is -1 (deleted embeddings)
-        filtered_results = [
-            (distances[0][i], indices[0][i])
-            for i in range(len(indices[0]))
-            if indices[0][i] != -1
-        ]
-        return [
-            (self.transform_similarity_score(dist, metric_type), int(idx))
-            for dist, idx in filtered_results
-        ]
-
-    def reset(self) -> None:
-        """
-        Reset the vector database to empty state.
-        """
-        if self.index is not None:
-            dim = self.index.d
-            self._init_vector_store(dim)
-        self.__next_embedding_id = 0
-
-    def _init_vector_store(self, embedding_dim: int):
-        """
-        Initialize the FAISS index with the given embedding dimension.
-
-        Args:
-            embedding_dim: The dimension of the embedding vectors.
-
-        Raises:
-            ValueError: If the similarity metric type is invalid.
-        """
-        metric_type = self.similarity_metric_type.value
-        match metric_type:
-            case "cosine":
-                faiss_metric = faiss.METRIC_INNER_PRODUCT
-            case "euclidean":
-                faiss_metric = faiss.METRIC_L2
-            case _:
-                raise ValueError(f"Invalid similarity metric type: {metric_type}")
-        self.index = faiss.index_factory(embedding_dim, "IDMap,Flat", faiss_metric)
-
-    def is_empty(self) -> bool:
-        """
-        Check if the vector database is empty.
-
-        Returns:
-            True if the database contains no embeddings, False otherwise.
-        """
-        return self.index.ntotal == 0
diff --git a/vcache/vcache_core/cache/vector_db/__init__.py b/vcache/vcache_core/cache/vector_db/__init__.py
new file mode 100644
index 0000000..fc736f2
--- /dev/null
+++ b/vcache/vcache_core/cache/vector_db/__init__.py
@@ -0,0 +1,15 @@
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
+    EmbeddingMetadataObj,
+)
+from vcache.vcache_core.cache.vector_db.strategies.hnsw_lib import HNSWLibVectorDB
+from vcache.vcache_core.cache.vector_db.vector_db import (
+    SimilarityMetricType,
+    VectorDB,
+)
+
+__all__ = [
+    "EmbeddingMetadataObj",
+    "VectorDB",
+    "HNSWLibVectorDB",
+    "SimilarityMetricType",
+]
diff --git a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_obj.py b/vcache/vcache_core/cache/vector_db/embedding_metadata_obj.py
similarity index 99%
rename from vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_obj.py
rename to vcache/vcache_core/cache/vector_db/embedding_metadata_obj.py
index 62914ad..2e45b82 100644
--- a/vcache/vcache_core/cache/embedding_store/embedding_metadata_storage/embedding_metadata_obj.py
+++ b/vcache/vcache_core/cache/vector_db/embedding_metadata_obj.py
@@ -11,8 +11,8 @@ class EmbeddingMetadataObj:
 
     def __init__(
         self,
-        embedding_id: int,
         response: str,
+        embedding_id: int = -1,
         prior: np.ndarray = None,
         posterior: np.ndarray = None,
         region_reject: List[str] = None,
diff --git a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/hnsw_lib.py b/vcache/vcache_core/cache/vector_db/strategies/hnsw_lib.py
similarity index 57%
rename from vcache/vcache_core/cache/embedding_store/vector_db/strategies/hnsw_lib.py
rename to vcache/vcache_core/cache/vector_db/strategies/hnsw_lib.py
index 7bb62c6..f641272 100644
--- a/vcache/vcache_core/cache/embedding_store/vector_db/strategies/hnsw_lib.py
+++ b/vcache/vcache_core/cache/vector_db/strategies/hnsw_lib.py
@@ -1,8 +1,11 @@
-from typing import List
+from typing import Dict, List
 
 import hnswlib
 
-from vcache.vcache_core.cache.embedding_store.vector_db.vector_db import (
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
+    EmbeddingMetadataObj,
+)
+from vcache.vcache_core.cache.vector_db.vector_db import (
     SimilarityMetricType,
     VectorDB,
 )
@@ -14,7 +17,7 @@
 
 class HNSWLibVectorDB(VectorDB):
     """
-    HNSWLib-based vector database implementation for efficient similarity search.
+    HNSWLib-based vector database implementation that stores both embeddings and metadata.
     """
 
     def __init__(
@@ -39,27 +42,36 @@ def __init__(
         self.M = None
         self.ef = None
         self.index = None
+        self.metadata_storage: Dict[int, EmbeddingMetadataObj] = {}
 
-    def add(self, embedding: List[float]) -> int:
+    def add(self, embedding: List[float], metadata: EmbeddingMetadataObj) -> int:
         """
-        Add an embedding vector to the database.
+        Add an embedding vector and its metadata to the database.
 
         Args:
             embedding: The embedding vector to add.
+            metadata: The metadata object associated with the embedding.
 
         Returns:
             The unique ID assigned to the added embedding.
         """
         if self.index is None:
             self._init_vector_store(len(embedding))
-        self.index.add_items(embedding, self.__next_embedding_id)
+
+        embedding_id = self.__next_embedding_id
+        self.index.add_items(embedding, embedding_id)
+
+        # Automatically set the embedding_id in the metadata
+        metadata.embedding_id = embedding_id
+        self.metadata_storage[embedding_id] = metadata
+
         self.embedding_count += 1
         self.__next_embedding_id += 1
-        return self.__next_embedding_id - 1
+        return embedding_id
 
     def remove(self, embedding_id: int) -> int:
         """
-        Remove an embedding from the database.
+        Remove an embedding and its metadata from the database.
 
         Args:
             embedding_id: The ID of the embedding to remove.
@@ -68,11 +80,15 @@ def remove(self, embedding_id: int) -> int:
             The ID of the removed embedding.
 
         Raises:
-            ValueError: If the index is not initialized.
+            ValueError: If the index is not initialized or embedding not found.
         """
         if self.index is None:
             raise ValueError("Index is not initialized")
+        if embedding_id not in self.metadata_storage:
+            raise ValueError(f"Embedding with ID {embedding_id} not found")
+
         self.index.mark_deleted(embedding_id)
+        del self.metadata_storage[embedding_id]
         self.embedding_count -= 1
         return embedding_id
 
@@ -98,7 +114,57 @@ def get_knn(self, embedding: List[float], k: int) -> List[tuple[float, int]]:
             self.transform_similarity_score(sim, metric_type) for sim in similarities[0]
         ]
         id_list = [int(id) for id in ids[0]]
-        return list(zip(similarity_scores, id_list))
+
+        # Filter out deleted embeddings (those not in metadata_storage)
+        results = []
+        for score, embedding_id in zip(similarity_scores, id_list):
+            if embedding_id in self.metadata_storage:
+                results.append((score, embedding_id))
+
+        return results
+
+    def get_metadata(self, embedding_id: int) -> EmbeddingMetadataObj:
+        """
+        Get metadata for a specific embedding.
+
+        Args:
+            embedding_id: The ID of the embedding to get metadata for.
+
+        Returns:
+            The metadata object for the embedding.
+        """
+        if embedding_id not in self.metadata_storage:
+            raise ValueError(f"Metadata for embedding ID {embedding_id} not found")
+        return self.metadata_storage[embedding_id]
+
+    def update_metadata(
+        self, embedding_id: int, metadata: EmbeddingMetadataObj
+    ) -> EmbeddingMetadataObj:
+        """
+        Update metadata for a specific embedding.
+
+        Args:
+            embedding_id: The ID of the embedding to update metadata for.
+            metadata: The new metadata object.
+
+        Returns:
+            The updated metadata object.
+        """
+        if embedding_id not in self.metadata_storage:
+            raise ValueError(f"Metadata for embedding ID {embedding_id} not found")
+
+        self.metadata_storage[embedding_id] = metadata
+        self.metadata_storage[embedding_id].embedding_id = embedding_id
+        return metadata
+
+    def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
+        """
+        Get all embedding metadata objects in the database.
+
+        Returns:
+            A list of all embedding metadata objects.
+        """
+        return list(self.metadata_storage.values())
 
     def reset(self) -> None:
         """
@@ -109,6 +175,7 @@ def reset(self) -> None:
         self._init_vector_store(self.dim)
         self.embedding_count = 0
         self.__next_embedding_id = 0
+        self.metadata_storage.clear()
 
     def _init_vector_store(self, embedding_dim: int):
         """
diff --git a/vcache/vcache_core/cache/embedding_store/vector_db/vector_db.py b/vcache/vcache_core/cache/vector_db/vector_db.py
similarity index 61%
rename from vcache/vcache_core/cache/embedding_store/vector_db/vector_db.py
rename to vcache/vcache_core/cache/vector_db/vector_db.py
index f28f8e2..6be68b4 100644
--- a/vcache/vcache_core/cache/embedding_store/vector_db/vector_db.py
+++ b/vcache/vcache_core/cache/vector_db/vector_db.py
@@ -2,6 +2,10 @@
 from enum import Enum
 from typing import List
 
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
+    EmbeddingMetadataObj,
+)
+
 
 class SimilarityMetricType(Enum):
     """
@@ -14,7 +18,7 @@ class SimilarityMetricType(Enum):
 
 class VectorDB(ABC):
     """
-    Abstract base class for vector databases.
+    Abstract base class for vector databases that store both embeddings and metadata.
     """
 
     def transform_similarity_score(
@@ -39,12 +43,13 @@ def transform_similarity_score(
                 raise ValueError(f"Invalid similarity metric type: {metric_type}")
 
     @abstractmethod
-    def add(self, embedding: List[float]) -> int:
+    def add(self, embedding: List[float], metadata: EmbeddingMetadataObj) -> int:
         """
-        Add an embedding to the vector database.
+        Add an embedding and its metadata to the vector database.
 
         Args:
             embedding: The embedding to add to the vector db.
+            metadata: The metadata object associated with the embedding.
 
         Returns:
             The id of the embedding.
@@ -54,7 +59,7 @@ def add(self, embedding: List[float]) -> int:
     @abstractmethod
     def remove(self, embedding_id: int) -> int:
         """
-        Remove an embedding from the vector database.
+        Remove an embedding and its metadata from the vector database.
 
         Args:
             embedding_id: The id of the embedding to remove.
@@ -78,6 +83,45 @@ def get_knn(self, embedding: List[float], k: int) -> List[tuple[float, int]]:
         """
         pass
 
+    @abstractmethod
+    def get_metadata(self, embedding_id: int) -> EmbeddingMetadataObj:
+        """
+        Get metadata for a specific embedding.
+
+        Args:
+            embedding_id: The id of the embedding to get the metadata for.
+
+        Returns:
+            The metadata of the embedding.
+        """
+        pass
+
+    @abstractmethod
+    def update_metadata(
+        self, embedding_id: int, metadata: EmbeddingMetadataObj
+    ) -> EmbeddingMetadataObj:
+        """
+        Update metadata for a specific embedding.
+
+        Args:
+            embedding_id: The id of the embedding to update the metadata for.
+            metadata: The metadata to update the embedding with.
+
+        Returns:
+            The updated metadata of the embedding.
+        """
+        pass
+
+    @abstractmethod
+    def get_all_embedding_metadata_objects(self) -> List[EmbeddingMetadataObj]:
+        """
+        Get all embedding metadata objects in the database.
+
+        Returns:
+            A list of all the embedding metadata objects in the database.
+        """
+        pass
+
     @abstractmethod
     def reset(self) -> None:
         """
diff --git a/vcache/vcache_policy/strategies/benchmark_iid_verified.py b/vcache/vcache_policy/strategies/benchmark_iid_verified.py
index 69827c1..9924932 100644
--- a/vcache/vcache_policy/strategies/benchmark_iid_verified.py
+++ b/vcache/vcache_policy/strategies/benchmark_iid_verified.py
@@ -7,10 +7,9 @@
 
 from vcache.config import VCacheConfig
 from vcache.vcache_core.cache.cache import Cache
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
     EmbeddingMetadataObj,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
 from vcache.vcache_core.similarity_evaluator import (
     SimilarityEvaluator,
     StringComparisonSimilarityEvaluator,
@@ -53,10 +52,7 @@ def setup(self, config: VCacheConfig):
         self.inference_engine = config.inference_engine
         self.cache = Cache(
             embedding_engine=config.embedding_engine,
-            embedding_store=EmbeddingStore(
-                embedding_metadata_storage=config.embedding_metadata_storage,
-                vector_db=config.vector_db,
-            ),
+            vector_db=config.vector_db,
             eviction_policy=config.eviction_policy,
         )
 
diff --git a/vcache/vcache_policy/strategies/benchmark_static.py b/vcache/vcache_policy/strategies/benchmark_static.py
index b192586..83e9876 100644
--- a/vcache/vcache_policy/strategies/benchmark_static.py
+++ b/vcache/vcache_policy/strategies/benchmark_static.py
@@ -4,7 +4,6 @@
 
 from vcache.config import VCacheConfig
 from vcache.vcache_core.cache.cache import Cache
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
 from vcache.vcache_policy.vcache_policy import VCachePolicy
 
 
@@ -40,10 +39,7 @@ def setup(self, config: VCacheConfig):
         self.inference_engine = config.inference_engine
         self.cache = Cache(
             embedding_engine=config.embedding_engine,
-            embedding_store=EmbeddingStore(
-                embedding_metadata_storage=config.embedding_metadata_storage,
-                vector_db=config.vector_db,
-            ),
+            vector_db=config.vector_db,
             eviction_policy=config.eviction_policy,
         )
 
diff --git a/vcache/vcache_policy/strategies/benchmark_verified_global.py b/vcache/vcache_policy/strategies/benchmark_verified_global.py
index df26f29..7d985ae 100644
--- a/vcache/vcache_policy/strategies/benchmark_verified_global.py
+++ b/vcache/vcache_policy/strategies/benchmark_verified_global.py
@@ -12,10 +12,9 @@
 from vcache.config import VCacheConfig
 from vcache.inference_engine import InferenceEngine
 from vcache.vcache_core.cache.cache import Cache
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
     EmbeddingMetadataObj,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
 from vcache.vcache_core.similarity_evaluator import SimilarityEvaluator
 from vcache.vcache_policy.vcache_policy import VCachePolicy
 
@@ -54,10 +53,7 @@ def setup(self, config: VCacheConfig):
         self.inference_engine = config.inference_engine
         self.cache = Cache(
             embedding_engine=config.embedding_engine,
-            embedding_store=EmbeddingStore(
-                embedding_metadata_storage=config.embedding_metadata_storage,
-                vector_db=config.vector_db,
-            ),
+            vector_db=config.vector_db,
             eviction_policy=config.eviction_policy,
         )
 
diff --git a/vcache/vcache_policy/strategies/verified.py b/vcache/vcache_policy/strategies/verified.py
index ffd93e0..ecf540f 100644
--- a/vcache/vcache_policy/strategies/verified.py
+++ b/vcache/vcache_policy/strategies/verified.py
@@ -16,13 +16,10 @@
 from vcache.config import VCacheConfig
 from vcache.inference_engine import InferenceEngine
 from vcache.vcache_core.cache.cache import Cache
-from vcache.vcache_core.cache.embedding_store.embedding_metadata_storage.embedding_metadata_obj import (
+from vcache.vcache_core.cache.vector_db.embedding_metadata_obj import (
     EmbeddingMetadataObj,
 )
-from vcache.vcache_core.cache.embedding_store.embedding_store import EmbeddingStore
-from vcache.vcache_core.similarity_evaluator import (
-    SimilarityEvaluator,
-)
+from vcache.vcache_core.similarity_evaluator import SimilarityEvaluator
 from vcache.vcache_policy.vcache_policy import VCachePolicy
 
 # Disable Hugging Face tokenizer parallelism to prevent deadlocks when using
@@ -144,10 +141,7 @@ def setup(self, config: VCacheConfig):
         self.inference_engine = config.inference_engine
         self.cache = Cache(
             embedding_engine=config.embedding_engine,
-            embedding_store=EmbeddingStore(
-                embedding_metadata_storage=config.embedding_metadata_storage,
-                vector_db=config.vector_db,
-            ),
+            vector_db=config.vector_db,
             eviction_policy=config.eviction_policy,
         )