BenMakesDataEasy @data_ben

I've created a benchmark that measures agent performance in basic Power BI report tasks, showing that directly modifying PBIR files is worse, more expensive, and (often) slower than using pbir-cli. The tests evaluate 30 simple tasks like "create a card that shows the total invoice lines, using " and a pass is when the agent can do it correct across 3/3 replicates. So if it does it 1 or 2 times, it still fails. The results were quite interesting. A plugin with skills and small CLIs for validation didn't meaningfully increase performance much (esp not for Opus 4.8), but ballooned token cost and time. In contrast, the pbir-cli alone was sufficient to increase performance significantly on those tasks, and adding the skill made it faster and take less tokens. "Dumber" models took longer with the CLI because they spent more time exploring or iterating, and seem worse at using skills in general. Obviously together with @maxanatsko I'm an author of the pbir-cli so it's appropriate to be skeptical of my benchmark that favors my own tool, but I did try to make these tests objective and fair. Note this benchmark only evaluates performance as a function of "can it complete the task" not "do the report / visuals look nice". I'm building a design evaluation as well but it takes a lot longer since design is subjective and coming up with good, repeatable tests that aren't handwavy marketing bs (like "make the report from this image") is challenging.

New! Round Range in Power BI 🚀 Works for columns, bars, scatters and lines :) June 2026 Update Soo many new small but useful features, I could definitely get used to that! 🤣✌🏻 #powerbi #update #roundrange #analytics

pbir-cli version 0.9.24 is out! Refresh or screenshot the Power BI Desktop canvas, support for thick reports and local model validation/queries, and some nice performance improvements. uv tool install --upgrade pbir-cli

I’ve quietly been working on this project for almost a year and I’m excited to finally share it! I wrote a book! This book is about what to do once you actually have the job. It’s called “Beyond the Interview: Building Your Data Career”. This is a really detailed and practical guide on how to grow your career quickly and efficiently. I share a lot of personal stories and give you behind the scenes into areas where I struggled and how I overcame them. I’m holding the Final draft before it goes live, hence the “not for resale”, but I’ll be launching it in about 1 week!

keyboard_arrow_left Previous keyboard_arrow_right Next

This article/video describes how to define optional parameters in DAX user-defined functions and set default values for parameters not specified by the caller. #dax #powerbi #udf sqlbi.com/articles/optio…

Super excited for the next release of the power-bi-agentic-development repo next week. It's all coming together!

You can subscribe to blog.tabulareditor.com to be notified when we publish new articles. We're planning a lot for our future content, including recording new videos on topics around Fabric/Power BI, analytics and agentic development. Our content motto: All signal, no noise! We also have a strict policy against AI-generated content, including AI-generated text and images. Everything we publish is written by a human and reviewed by humans; AI is only enabling creation of new deterministic tools that we use to create our content.

At Power BI DC days, Shreya Harvu and Bhargav Patel remind us why good semantic models are are still critical if you want AI to interact with your data.

A preview of the next version of the pbir-cli, which can refresh and screenshot the Power BI canvas using the new Power BI desktop preview feature from Microsoft. This makes the agent a lot more powerful, and the experience more fun / satisfying. View in full-screen!

A new version of the pbir-cli / power-bi-agentic-development skills let you refresh the Power BI canvas. With the te-cli & chaining commands you (or an agent) can do cool stuff, like fixing reports after model renames. Will demo this in our webinar later today.

Join me at Power BI Days D.C. for learning, discussion, and probably a little chaos.

I am glad to see new videos describing specific Synoptic Panel features because it means I didn't have to record them! This is a tutorial of the Map Selector feature: you create a DAX measure that returns the name of the map to display. This way, you can respond to the filter context and make custom choices based on selections in other visuals.  youtube.com/watch?v=nnDcOq…

Tomorrow we're giving a webinar about the te-cli. Sign up here free: tabulareditor.com/resources/upco… I'll speak briefly about why CLIs are interesting for agents, even if you favor other approaches like an MCP server.

added more tests (merge, append, safe append etc) considering it is using github runner with modest hardware, that's not bad at all github.com/djouallah/duck… #deltalake #rust #arrow #duckdb

@awsaxton Very excited for you!

@kurtbuhler that's pretty cool : )

Added a detection which estimates and warns whether a session or week will exceed its limits should usage continue linearly at the same rate, and throws a blinking warning if so. The bar is a linear gauge in 20% increments which fills in layers to warn of usage.

Kurt Buhler @kurtbuhler

3 weeks ago

I've formed a habit of changing one thing in my status line every time I update my memory files when wrapping up my work for the day. It's such a fun little ritual.

2 0 13 5K 11

I rebuilt #duckrun as a thin #dbt adapter on top of the excellent dbt-duckdb: data is exchanged lazily via the Arrow C-stream interface and materialized as real Delta Lake tables with #delta-rs. #duckdb #deltalake github.com/djouallah/duck…

If you're a "power bi person" I encourage you to not see data apps as a threat to your work or preferred tooling. I really think we should embrace this as something that opens a whole bunch more possibilities. Don't think in terms of "replace/displace" - just try it.

Kurt Buhler @kurtbuhler

3 weeks ago

Some key differences when comparing data apps vs reports if you want to decide where to visualize your data. Boils down to - if you need a nice dashboard with the best visuals -- fabric apps. If you need basic reporting -- power bi. Some nuance in the middle, as always.

2 2 26 4K 18

My Parkbench CLI tool allows you to practice streaming events to DuckLake so you can see how it responds. When streaming data the table updates but no Parquets are written. This is DuckLakes's inlining feature in action. See the tool here: github.com/early-signal-t…