Comments for bicortex

Comment on Building Snowflake-Centric GenAI Data Apps: Why SPCS Beats Streamlit and External Hosting by admin

admin — Tue, 11 Aug 2026 06:36:29 +0000

In reply to MJ.

I’m not sure I fully agree here; even with a few years of integrating AI into these apps e.g. Power BI – Copilot, Tableau – Einstein, Qlik – Insight Advisor etc. they are still brittle and unsophisticated when it comes to creating natural language and conversational analytics experiences. In addition, you can’t easily whitelist your app, customize its UI/UX and pick a model you like, not to mention all these applications require cross-boundary data movement. I think we still have a long way to go but having options is a good thing, right 🙂

Comment on Building Snowflake-Centric GenAI Data Apps: Why SPCS Beats Streamlit and External Hosting by MJ

MJ — Tue, 11 Aug 2026 04:24:57 +0000

Nice demo but you can probably wait a few more months and get most of this functionality in Power BI or Qlik, with all the security built-in.

Comment on Building Snowflake-Centric GenAI Data Apps: Why SPCS Beats Streamlit and External Hosting by admin

admin — Wed, 05 Aug 2026 23:19:42 +0000

In reply to Freddy L. Any model hosted by Snowflake or accessed through Cortex Agents or Cortex Complete can be used for this solution. As per Snowflake documentation: "Cortex Agents supports models from the Claude (Anthropic), GPT (OpenAI), Grok (SpaceX) and Gemini (Google) families"

Comment on Building Snowflake-Centric GenAI Data Apps: Why SPCS Beats Streamlit and External Hosting by Freddy L

Freddy L — Wed, 05 Aug 2026 08:21:13 +0000

Hi Martin

This demo looks interesting, what LLM can you use with this solution, is it only Snowflake Arctic or can Anthropic, OpenAI or other providers’ models also be used?

Comment on Protecting PHI & PII data at Scale with LLMs using Snowflake, LangGraph and Streamlit for Human-in-the-Loop QA by Leo Mattias

Leo Mattias — Mon, 25 May 2026 05:53:33 +0000

Hello

Kind of what I’ve been looking for – thanks for posting. I don’t use Snowflake but gives me an idea on how I can use this approach to do something similar in Azure.

L.M.

Comment on Protecting PHI & PII data at Scale with LLMs using Snowflake, LangGraph and Streamlit for Human-in-the-Loop QA by admin

admin — Mon, 18 May 2026 08:48:04 +0000

In reply to Do Ho. Ahh, yes, the Streamlit app, as implemented in the current codebase, is using Snowflake for inference but is designed for local runtime in your local Python venv. So the deploy_spcs script does not include Streamlit being published to Snowflake. However, it would not be a big effort to re-platform it so it either runs on a second SPCS service or as a Native App Framework project.

Comment on Protecting PHI & PII data at Scale with LLMs using Snowflake, LangGraph and Streamlit for Human-in-the-Loop QA by Do Ho

Do Ho — Fri, 15 May 2026 17:24:24 +0000

Hi Martin

I looked at the code you provided but it seems the Streamlit app can only run on a local environment. Can you confirm the deploy_spcs script deploys the QA app to a Snowflake-hosted environment?

Thank you.

Comment on Snowflake Scale-Out Metadata-Driven Ingestion Framework (Snowpark, JDBC, Python) by José Fernández

José Fernández — Sun, 01 Feb 2026 15:56:06 +0000

Great post Martin and I found that the source code worked well. My company was about procure and deploy an expensive ETL product from a well know vendor to tackle data integration (we’re quite new to Snowflake). But since most of our data sources are relational DBs, and your approach does not require additional tools, we managed to save a lot of $$$. Thanks again.

Comment on Snowflake Scale-Out Metadata-Driven Ingestion Framework (Snowpark, JDBC, Python) by admin

admin — Wed, 07 Jan 2026 12:45:26 +0000

In reply to Huo Wang. The answer is definitely yes, providing you find a compatible JDBC driver and adjust catalog queries for metadata extraction to correspond to the RDBMS you're working with.

Comment on Snowflake Scale-Out Metadata-Driven Ingestion Framework (Snowpark, JDBC, Python) by Huo Wang

Huo Wang — Wed, 07 Jan 2026 10:42:23 +0000

Nice post Martin. I was wondering if this can be used with other database vendors like MySQL, but since it’s using JDBC under the hood, I guess the answer is yes?