Hey, nice to meet you 👋 Let's build awesome Data Science demos to get you noticed! Follow me ✔️ 🎈 For news: The latest on @streamlit & Data ecosystem 🐍 For tips: Python for Showcasing Data Analysis & Automating Boring Stuff 🎞️ For fun: Data & Tech Video skits Examples ⬇️

i love writing. i never let agents write for me. whether it's docs, or tweets, or shitposts, or even prompts (okay maybe that one's harder as agents manage other agents) there's a simple reason for that: i still strongly believe that writing is thinking. it's a skill i hope to never give up. it's also important to me that whatever i say is always authentic, mistakes and all. my favorite posts are the ones that i just full send without any editing i suspect it also helps with prompting. being able to clearly express what i want, and knowing the vocabulary i need to be able to direct my agents does seem to make for more effective usage of agents. in the same way that engineers have an advantage when we prompt agents to write code, because we are intimately familiar with the words that make up the discipline knowing the right word incantations to get an agent to produce a desired goal almost feels like spellcasting. as models get more capable, and that input box looks more and more like a magic box that can output anything, words become the interface between you and the genie there may come a time when we give up the keyboard in exchange for brain-computer interfaces. but until then, i'm writing every word.

How to effectively debug with agents when you no longer read or understand the code In this post, I propose a debugging framework for agents that doesn’t rely on human knowledge of code to get to the right “root cause” The TL;DR are the screenshots. Link to post in the reply.

get off the TL sometimes, it’s good for you, brother they’re still writing java 8 out there

🥳 Excited to start revealing what we've been working on in the last few months. First, we decided to reinvent Kubernetes for agentic workloads with statefulness and fast resumption. Secondly, we are building an agentic orchestrator that will be Google's open agentic orchestrator and runtime. github.com/google/ax

I'm having a lot of fun with @OpenWebUI -> OWUI Computer -> @pidotdev -> @OpenRouter -> @Zai_org GLM Flash / @deepseek_ai V4 Flash on @Tailscale All I'm missing is a machine strong enough to host some model through @huggingface and I would be 100% satisfied with my 24/7 stack

I notice I stopped using plan mode (replaced by @mattpocockuk grill-me) or MCP (agents are good at using authenticated CLI clients), so I tried out @pidotdev which doesn’t include either by default, yet makes it easy to integrate any other extension I get the interest now 👀

X: Astra does my video editing/blog writing/3D modeling tasks now! Me: My WiFi is broken, it only displays "Can't connect to server", I have no clue what is happening, go through the Windows logs and fix plz 🥹 (apparently some 802.11ax setting on an older router)

As someone who poured 2 years of my life into learning Ableton Live to produce music before being a Youtuber... This is the best demo of Astra I've ever seen, song is at the end 😏

Super happy to officially announce that we are a̶c̶q̶u̶i̶r̶i̶n̶g̶ joining forces with @nvidia 🔥 Here is a more personal take: AI is at an inflection point. Open source AI can become less relevant in the coming years if the big closed labs run away with it, OR it can become the foundational fabric of the next phase of human civilization. Those are vastly different outcomes, and we need the critical mass to ensure we give our collective best shot to the second outcome. Given @JensenHuang's stance on open source AI and how he stepped up to defend it when it was under threat earlier in the summer, NVIDIA was the only partner we truly considered. HF will remain an independently run, neutral platform. This gives fuel to our long-term vision and mission of unlocking the community's progress to ensure that AI, which is the greatest breakthrough of our lifetime, is accessible to as many people as possible.

Not enough people realize that if you can vibe a framework or app over a weekend, there is no reason I can't also do that. What you want is somebody who vibes & invests a significant amount of time to solving a specific problem over the long-term.

If you want to showcase your senior engineering skills, maybe start a Youtube channel doing /grill-me sessions using @mattpocockuk 's skill over a semi-large codebase you know nothing about The way you react and even pushback on those demanding questions would be valuable 🤔

🌸 Youtube watching hours to get partnered doubled from 4k to 8k This may push me to do more 1h unedited talking head videos than 10 minutes overedited code explanations 🤔 and move more of my editing into Shorts Look forward to more of my yapping on Youtube 👀

🌸 DataFan, every year, I get the "this technology will replace Excel" issue Everytime top executives ask for the Excel file to play with the data Excel is so ingrained in every decision making process, that I just brace myself for an Excel export of every data tool I create

🌸 Streamlit 1.59 release from a month ago. Better late than never! --- IMO We're seeing more and more dataframe related changes. I wonder if Streamlit is planning to make their Dataframe one of the best Python interactive web spreadsheets editable by coding agents 👀

🌸 Streamlit theme for font, color and spacing configuration is defined inside .streamlit/config.toml Coding agents with Streamlit agent skills have an easy time editing it with inspiration from external references! Code: github.com/andfanilo/streaml…

🌸 Streamlit 1.60.0 is out! (I'm back to "build it in Illustrator in 1 hour" for those carousels, going against the GenAI HTML slides trend. Layout, font and colors in progress 👀)