Release review

CHANGELOG.md

@@ -3,8 +3,6 @@
 ## Bug Fixes (requires `--full-refresh`)
 - Applied `coalesce_cast` macro to all relevant fields that are being concatenated into `comment_markdown`, as any concatenation in Snowflake with a null value returns null. We coalesced 'UNKNOWN' on a string field, and '1970-01-01 00:00:00' on a timestamp field to ensure Snowflake returns chunks of texts for all comments with null components.
 - Fields are now coalesced in these intermediate models:
-
-```yml
   - Hubspot
     - `int_rag_hubspot__deal_comment_document`: 
       - `email_title` (string) 
@@ -27,12 +25,10 @@
     - `int_rag_zendesk__ticket_document`:
       - `title` (string) 
       - `created_on` (timestamp)
-```
 
 - Corrected syntax errors for the `default_variable` in `stg_rag_hubspot__engagement_email` and `stg_rag_hubspot__engagement_note`.
 - Updated joins to ensure `engagement_deal` is the base in the `int_rag_hubspot__deal_comment_document` CTEs.
-- Added `most_recent_document` CTEs to `int_rag_*__deal_comment_documents_grouped` models in Hubspot, Jira and Zendesk to ensure we only bring in the `most_recent_chunk_update` by document. 
-  - The initial group by did not work properly, as `chunk_index` and would always be different for the `deal_id` on the group by statement and lead to different chunk time updates for each index.
+- Added `most_recent_document` CTE to `int_rag_*__deal_comment_documents_grouped` models in Hubspot, Jira and Zendesk to correctly bring in the `most_recent_chunk_update` by document.
 - Brought in `engagement_type` from the Hubspot `engagement_deal` source to produce proper chunk records in the `rag__unified_document`. 
 - Added filters on `email` and `note` types in `int_rag_hubspot__deal_comment_document` when creating email and note chunk records.