568 - Fabric ontology launched with two frustrations. Building anything in that front end was painful, and nobody could say what to do with the result. An agent just took over the manual part.
@powerbitips.bsky.social
@powerbitips.bsky.social568 - Fabric ontology launched with two frustrations. Building anything in that front end was painful, and nobody could say what to do with the result. An agent just took over the manual part.
@powerbitips.bsky.social568 - Fabric ontology isn't a green grass plain anymore. There are clay dirt roads starting to form, and they don't lead anywhere yet. The marks point somewhere.
@kyisaiah47.thecompound.techVirtual warehouses run queries as independent compute resources. You can resize compute while data stays exactly where it is. #VirtualWarehouses #DataEngineering
@attlogics.bsky.socialNot sure which Databricks certification to choose? Start by looking at your skills, experience, and career goals. The right choice can help you plan your learning journey with more clarity.
Learn more: www.atingupt.in
@powerbitips.bsky.social567 - Customers need a portal to drop files into. Now you can build the uploader as a Fabric app and the file lands in your lakehouse on its own. Here's what that unlocks for notebooks.
@infoq.comWhat is an ontology, and why is #Netflix using one for observability?
An ontology encodes meaning through simple Subject-Predicate-Object triples, helping connect data and make it more useful for enterprise AI.
🎥 Watch the full video: buff.ly/fIoqHqD
@quansight.comZero or null. Mathematically correct or practically useful. Marco Gorelli digs into a real design trade-off between Polars and SQL, and shares an opinion on which one gets it right.
Full talk: buff.ly/L0duaRh
@powerbitips.bsky.socialNot just asking for a notebook. Tasking an agent to go research a problem and figure out your slowly changing dimensions. Data engineering is due for a big shakeup.
@powerbitips.bsky.socialAn OS ships with a write up the issue skill and it lands differently. Data pipelines need that. Semantic models need that. Users hit a gap and something should write up what's missing.
@powerbitips.bsky.socialA brand new external data source cracks the whole thing open. Live or staged, how you authenticate, whether it's a secret. More moves than a chess game, and plenty of them are bad.
@quansight.comSupporting multiple dataframe backends usually means maintaining multiple codebases. Narwhals solves that with zero dependencies, a stable API, and full static typing. Marco Gorelli explains the design.
Watch the full talk from PyData London: buff.ly/9f9rQEI
#Narwhals #Python #DataEngineering
@powerbitips.bsky.social562 - AI can write 10,000 lines of net new code in two minutes. Great. Now prove it's correct. The bottleneck moved somewhere new, and most teams haven't solved it yet.
@infoq.comThe real #AIagent problem? Messy enterprise semantics.
If Marketing and Finance can’t agree on “active customer,” how can an AI agent?
Models aren’t the problem. Conflicting definitions are!
🔗 Watch Fabiane Nardon’s full talk: buff.ly/TAbVpH4
@powerbitips.bsky.social561 - Snowflake keeps getting easier to work with inside Power BI. Databricks keeps grinding against it. The friction is pushing teams into a split that changes who owns which half of the stack.
@powerbitips.bsky.social561 - A client asked for a data warehouse a couple years back. Nothing was in SQL. It was all SharePoint lists. That conversation had to go somewhere before the word warehouse came up again.
@powerbitips.bsky.social561 - The best practice articles for the warehouse showed up the same week the cost went up. Call it a coincidence if you want. The read on why both landed together is less generous than that.
@powerbitips.bsky.social560 - Thirty minutes. Data landed in a lakehouse, semantic model built, full report designed, and Power BI Desktop never got opened once. The theme it picked out on its own is the part that got me.
@quansight.comEveryone rushes to compare Polars and SQL on speed and syntax. Marco Gorelli points out that almost nobody is comparing the semantics, the actual mental models behind each one.
Full talk from PyData London: buff.ly/uInE29g
#Polars #SQL #PythonData #DataEngineering #Quansight