Snowflake Connection for AIFAQ
Team Member | Goal |
| |
|---|---|---|---|
Bobbi Muscara | Organize group communications. |
| |
Madugula Jayaram | Implement the snowflake connection pipeline for AIFAQ |
| |
Sumana |
|
|
|
Building RAG Applications with Cortex AI
Mentorship Applicants
completed application |
|
|
|---|---|---|
Altafur Rahman |
|
|
Madugula Jayaram |
|
|
Siddharth Chauhan |
|
|
Tushar Chandra |
|
|
Technical Goals:
Data Pipeline Integration: Build a robust connector to extract structured FAQ or knowledge base data from Snowflake tables.
Authentication & Security: Implement secure authentication (OAuth, token-based) for enterprise Snowflake accounts.
Connector API: Develop a simple API interface to trigger data pulls, syncs, and refreshes on-demand.
Auto-Sync and Scheduling: Allow scheduled pull jobs and delta updates (new or modified rows only).
Business/Strategic Deliverables:
Enterprise Use Case Demo: Showcase pulling an enterprise customer’s knowledge base into AIFAQ using Snowflake in a demo environment.
Documentation & SDK: Release an open-source SDK or connector package on GitHub with easy-to-follow examples.
Outbound Messaging: Write 1 LinkedIn article or case study targeting Snowflake’s huge enterprise customer base (joint marketing potential if successful).
NOTES
Himanshu Mahajan
09:36 AM
HM
i am too familiar with this workflow it looks exactly like the Workflow we have in our product at my current startup job
The system uses Snowflake to securely ingest, store, and update documents and their embeddings, applies vector search with Cortex for fast retrieval, enforces user access control with RBAC, and dynamically feeds relevant content to LLMs for context-aware responses, all orchestrated automatically with Snoopy.
connecting the data to our chatbot using this snowflake connectors
We can also dynamically update our database knowledge by extracting knowledge in real time using API hits
1
MADUGULA JAYARAM
09:46 AM
MJ
custom tools of snowflake can be easily integrated to our chatbot so that we can make our chatbot gets data from the customer and all , securely with privacy .
Ayush Jha
09:52 AM
AJ
For hallucination reduction, we can go with reranking and filtration too and can change our search algo for better semantic capturance