I am passionate about creating products and bringing them to market. AI helps accelerate that journey.
- Nvidia / LlamaIndex Developer Contest https://wgreunke.github.io/nvidia_llama_hackathon/
- Visualize events from different news articles related to Hurricane Helene on a map. Built with React / LlamaIndex / Nvida NIM - abacusai/dracarys-llama-3.1-70b-instruct LLM.
- AWS AI Hackathon - Second Place: Sugar Reduction App / Osebusters that uses multi-modal llm to parse nutrition labels on food to show the amount of sugar per serving as well as the steps you have to walk to work off the calories per serving.
- Tidb Hackathon: Fact Checker Allows you to determine if a statement is true or false by comparing the statement to reference documents held in a vector databases. A RAG application built using Tidb vector database, SentenceTransformers vector embedding, LangChain and GPT-4o mini.
- RAG: Gutenberg Library AI Query Tool A Retrieval Augmented Generation(RAG) tool that allows you to query books from the Gutenberg library that were not included in the ChatGPT-3.5-Turbo model. Built using Python and deployed with Streamlit. Code
- Chatbot Front End A simplified front end for generative chatbots.
- No Code/Text Input tool that allows users to generate Javascript games using only text input. Built with Python and Streamlit.
- Image Classification using Convoluted Neural Network(CNN) fine-tuned with TensorFlow
- Introduction to AI Chatbots
- How to query SQL data using AI
- How to explain neural networks to a 5 year old
- Difference between RAG and Fine Tuning
- GreekGraph - An option trading strategy visualization tool.
- Pizza and Beer - Problem: After windsurf racing every Tuesday night in Foster City, the whole group goes out for pizza and beer. Splitting the bill is hard because not everyone drinks beer. I created an app that calculates how much each person owes at the end of the night in Python and Streamlit. You can see the code here: GitHub