Before a government agency puts an AI agent in front of the public, one question always comes first: is it safe? Gatehouse is my answer to one part of that question. It serves GovTech Singapore's LionGuard safety model inside a Databricks workspace, ...
Why TetrisI wanted to know if one Databricks function can play Tetris, and what Unity Gateway sees when an AI agent hammers it move after move.It started with the ai_decide announcement by Jane Zhang, Matthew Ding and Archika Dogra. Their demo is Sna...
A document pipeline can fail in a way that looks a lot like success. The job finishes, the index status is green, the file count matches what you uploaded. Then someone asks a question and gets nothing back. No error, just an empty answer.It's a comm...
Lakeshift is a Databricks App that I built for one job: moving legacy SQL and ETL code to Databricks in a way that a team can measure, check and trust. It sizes the code first, makes a rule-based draft with Databricks Labs Lakebridge, lets two AI mod...
Hi everyone,Part 2 of my Helsinki live transit project is out. In part 1 I built a Databricks App that follows every tram and metro train in Helsinki in real time: a Lakeflow pipeline on HSL's live MQTT feed, Lakebase behind the live map, and a Genie...
Hey @SumeshKashyap Thanks for writing this up, and for the question at the end. I found it interesting enough that I ran an experiment on it.I set up bike, walk or wait inputs like the ones your agent gets back from its tools (16 scenarios over six H...
I ran the same 24 coding tasks through three arms: fixed Haiku, fixed Opus and Databricks-managed selection. The managed-selection arm passed all 24 tasks. Its 24 routing records also said task profiling was unavailable.That is not a contradiction. T...
Suppose a RAG application returns this:{ "response": "You can return the item within 90 days.", "citations": ["returns-policy"] }The cited document allows returns within 30 days.A citation-presence check should pass this response. A check that th...
The problem: rethinking routing in multi-agent systems on DatabricksDatabricks native supervisor agent pattern is a solid default for orchestrating multiple Genie spaces and tools  but it comes with a routing constraint worth examining. The superviso...
Data Engineering 1.0 was largely about integrating, moving, consolidating massive data to support enterprise decisions. Enterprises built ETL pipelines, data warehouses, data lakes and orchestration platforms to bring data together reliably. Data Eng...
Every sales manager I've ever talked to has the same quiet frustration: the CRM has all the answers, but getting to them takes forever. You want to know which deals are actually at risk this week, and instead you're pivoting spreadsheets, cross-refer...
I just earned my Databricks Certified Generative AI Engineer Associate Certification, and in this post, I’m sharing the key tips, resources, and including what confused me, what actually worked, and the traps I nearly fell into. Why I Took This ExamI...
Thanks for sharing,do you know what is the passing score to clear the exam?
How AI-powered development accelerated my data engineering workflow Watch the Complete Development Process YouTube Video: See the entire 30-minute development sessionThis is a screen recording without voice narration showing the complete development ...
This is super insightful @Pat, thanks for sharing this with the Community!
