fix one link

docs/website/docs/reference/explainers/how-dlt-works.md

@@ -39,7 +39,7 @@ During the normalization phase, `dlt` inspects and normalizes your data and comp
 
 ### Load
 
-During the loading phase, `dlt` first runs schema migrations as needed on your destination and then loads your data into the destination. `dlt` will load your data in smaller chunks called load jobs to be able to parallelize large loads. If the connection to the destination fails, it is safe to rerun the pipeline, and `dlt` will continue to load all load jobs from the current load package. `dlt` will also create special tables that store the internal dlt schema, information about all load packages, and some state information which, among other things, are used by the incrementals to be able to restore the incremental state from a previous run to another machine. Some ways to control the loading phase are by using different [`write_dispositions`](../../general-usage/incremental-loading#choosing-a-write-disposition) to replace the data in the destination, simply append to it, or merge on certain merge keys that you can configure per table. For some destinations, you can use a remote staging dataset on a bucket provider, and `dlt` even supports modern open table formats like [deltables and iceberg](../dlt-ecosystem/destinations/delta-iceberg), and [reverse ETL](../../dlt-ecosystem/destinations/destination) is also possible.
+During the loading phase, `dlt` first runs schema migrations as needed on your destination and then loads your data into the destination. `dlt` will load your data in smaller chunks called load jobs to be able to parallelize large loads. If the connection to the destination fails, it is safe to rerun the pipeline, and `dlt` will continue to load all load jobs from the current load package. `dlt` will also create special tables that store the internal dlt schema, information about all load packages, and some state information which, among other things, are used by the incrementals to be able to restore the incremental state from a previous run to another machine. Some ways to control the loading phase are by using different [`write_dispositions`](../../general-usage/incremental-loading#choosing-a-write-disposition) to replace the data in the destination, simply append to it, or merge on certain merge keys that you can configure per table. For some destinations, you can use a remote staging dataset on a bucket provider, and `dlt` even supports modern open table formats like [deltables and iceberg](../../dlt-ecosystem/destinations/delta-iceberg), and [reverse ETL](../../dlt-ecosystem/destinations/destination) is also possible.
 
 ## Other notable features