favs.blue
Sign in
@davidgasquez.com avatar

David Gasquez

@davidgasquez.com
Most liked Most reposted
All time Last month
  1. 1

    Wrote about why the AT Protocol is the ultimate API for you and your agents!

    davidgasquez.com/atproto-agents

    We need more people building on the atmosphere!

    My atmosphere data lives at at://davidgasquez.com. You can explore it all without API keys, auth, or arbitrary HTML to parse. Your agent can browse it, query it, and link to it. Getting someone latest posts is one click/curl away. That’s the pitch. The rest of this post is why it works, and why it might be the substrate your agents have been waiting for. Hostile Platforms You’ve probably experienced more than once recently, platforms being hostile to you (or your agent) getting the data out: throttled or limited APIs, bans for scraping, no identity persistence, no real-time access, walled gardens. Every integration is a deal that can be revoked when the CEO wakes up in a bad mood. The AT Protocol is an amazing technology that I think is very underrated for agents! Turns out, the properties Bluesky (the biggest AT Protocol application as of today) needed for humans (portable identity, open data, federated infrastructure, structured schemas) happen to be exactly what you’d want for agents too. As a quick and dirty intro (you should read the official ones, or this great walkthrough by mackuba), in the atproto world, every user is a personal, signed JSON repository. Every record (a post, a like, a follow, a photo, a blog post, anything really) is addressable by an at:// URI, has a public schema (a Lexicon), and is broadcast in real time on a global event stream (the firehose). Dan Abramov explains all of this in more detail and clarity in “A Social Filesystem”.
    125 15
    May 18, 2026
  2. 2

    New post up! ✨

    Exploring AT Protocol with Python to visualize the #databs social graph!

    davidgasquez.com/exploring-at...

    Took less than 1 hour to get the data and plot it. Amazing what you can do with open APIs and great SDKs!

    a gr
    108 20
    Nov 14, 2024
  3. 3

    It doesn't only do Office files! You can use it to get YouTube transcript or transform articles to Markdown!

    uvx --python 3.12 markitdown "https://www.youtube.com/watch?v=6KcV1C1Ui5s" | llm "Summarize the video"

    kepano @stephango.com
    Microsoft just released a tool that lets you convert Office files to Markdown. Never thought I'd see the day. Google also added Markdown export to Google Docs a few months ago. github.com/microsoft/markitdown
    GitHub - microsoft/markitdown: Python tool for converting files and office documents to Markdown.
    Python tool for converting files and office documents to Markdown. - microsoft/markitdown
    73 9
    Dec 15, 2024
  4. 4

    New data oriented file format just dropped.

    FastLanes, "like Parquet, but with 40% better compression and 40× faster decoding". 👀

    Seems it can exploit correlations between columns and have fully SIMD friendly encodings to help with vectorization.

    github.com/cwida/FastLa...

    GitHub - cwida/FastLanes: Next-Gen Big Data File Format
    Next-Gen Big Data File Format. Contribute to cwida/FastLanes development by creating an account on GitHub.
    70 15
    Jul 24, 2025
  5. 5

    Inspired by @simonwillison.net llm-docs repo, I did a similar one compiling projects docs into single TXT files that can be fed to LLMs.

    Right now, it only has atproto docs but already been useful to me to answer random questions about the project.

    github.com/davidgasquez...

    GitHub - davidgasquez/docs-to-llmstxt: 🤖 Compile docs into text files for LLMs
    🤖 Compile docs into text files for LLMs. Contribute to davidgasquez/docs-to-llmstxt development by creating an account on GitHub.
    64 6
    Nov 19, 2024
  6. 6

    Wrote a small post exploring different approaches to one of the most common tasks I perform as a data engineer.

    Getting data in batches from random APIs.

    davidgasquez.com/async-batch-...

    David Gasquez
    Posts, notes, and projects by David Gasquez
    37 5
    Nov 12, 2024
  7. 7

    Wonderful post by @summerscope.bsky.social! Hat tip to @mariozechner.at.

    The Human-in-the-Loop is Tired.

    pydantic.dev/articles/the...

    The Human-in-the-Loop is Tired
    On reward functions, dopamine, and what it actually feels like when the code starts writing itself
    34 6
    Jul 16, 2026
  8. 8

    The future is now!

    - Dial keys, not IPs, thanks to Iroh
    - Fetch hashes, not URLs, thanks to DASL
    - Build on shared records, not siloed databases, thanks to the AT Protocol

    Iroh @iroh.computer
    We just released the first stable version of iroh, v1.0! Read about it: www.iroh.computer/blog/v1 We're super proud of where we landed with iroh's 1.0. Since version 0.35 it feels like we've entirely replaced the "engine underneath". Are you a fellow key-dialer? Come tell us how you use iroh!
    Iroh 1.0 - Dial Keys, not IPs
    Iroh 1.0 is out. Now is the time to build.
    30 6
    Jun 16, 2026
  9. 9

    TIL about `uv tool install`! 🤦

    Previously, I was aliasing packages to `uvx package` and dealing with small lag (installing deps from cache). This is the perfect way to manage Python CLI tools.

    Installed llm, open-interpreter, and files-to-prompt.

    29 4
    Nov 29, 2024
  10. 10

    Wrote about "Barefoot Data Platforms", my take on minimal and agent friendly data platforms!

    davidgasquez.com/barefoot-dat...

    So, how does a minimal and opinionated data platform look like? Minimalism for me here means low abstraction and no frameworks. You write your scripts. These scripts are assets, and these assets depend on each other 1. That’s it. The opinion side of things is the interesting part. Here are the key ideas of the Barefoot Data Platform I built: Classic functional, declarative, independent, composable, and idempotent transformations Co-located assets, metadata, and documentation so assets have everything needed to be understood and reproduced in the same file One storage layer. DuckDB or SQLite, Parquet files, or whatever you want as long as you define a materialize function Many automated checks (ruff with all the rules enabled, ty, data tests, prek hooks, custom bdp check, …) for the agents
    26 2
    Feb 3, 2026
  11. 11

    Write a small and rough post on learnings and the process I took to export 5000 datasets (200GB) from the Spanish National Statistic Institute into 500MB worth of Parquet files on HuggingFace.

    davidgasquez.com/exporting-in...

    David Gasquez
    Posts, notes, and projects by David Gasquez
    David Gasquez @davidgasquez.com
    Worked on a small side-project during the weekend to export ALL tables from The Spanish National Statistics Institute. - Code: github.com/davidgasquez... - HuggingFace Dataset: huggingface.co/datasets/dav...
    26 4
    Nov 4, 2024
  12. 12

    Added @duckdb.org and @evidence.dev docs to the docs-to-llmstxt project. It will generate 1 file per project with all their docs for easier LLM feeding!

    github.com/davidgasquez...

    docs-to-llmstxt/docs at main · davidgasquez/docs-to-llmstxt
    🤖 Compile docs into text files for LLMs. Contribute to davidgasquez/docs-to-llmstxt development by creating an account on GitHub.
    25 2
    Nov 20, 2024
  13. 13

    Had a lot of fun teaching Cursor Agent new tricks. ✨

    - How to search the web from the CLI
    - How to run any Python / JavaScript code without installing dependencies
    - How to read links as Markdown

    davidgasquez.com/cursor-agent...

    Very cool seeing it putting everything together when working!

    24 1
    Jan 9, 2025
  14. 14

    Behold is the Bluesky World!

    Got "The Bluesky Dataset" from @andrewconner.com and used Graphext (from @victoriano.bsky.social) to visualize all the connections. It looks beautiful! 😍

    The best thing is that you can explore it too! 👇

    https://public.graphext.com/fe28f76fe67518ab/index.html

    24 7
    Apr 23, 2023
  15. 15

    DuckDB for processing, Parquet for publishing. Simple and powerful!

    www.robinlinacre.com/parquet_api/

    Why parquet files are my preferred API for bulk open data
    Open data should be served as CORS-enabled parquet files rather than using a custom API
    Robin Linacre @robinlinacre.bsky.social
    New blog: Why DuckDB is my first choice for data processing: www.robinlinacre.com/recommend_du...
    Why DuckDB is my first choice for data processing
    Why DuckDB has become my go-to tool for data processing, offering simplicity, speed, and powerful features.
    23 3
    Mar 17, 2025
  16. 16

    Wrote a post about debugging @dagster.io in VSCode / Cursor!

    If you're like me and use a lot of notebooks to write code, this might be a great alternative that relies on VSCode / Cursor debugger for a smooth development experience with Dagster or similar libraries.

    davidgasquez.com/debug-dagste...

    21 4
    Dec 28, 2024
  17. 17

    Databases are a wonderful rabbit hole to learn about many awesome data structures, OS instructions, network protocols, algorithms, and other random things that might help you understand a bit better what's going on with computers. 😅

    philip @heltweg.org
    Just read this well written article on database fundamentals, maybe interesting for someone else as well: tontinton.com/posts/databa... :)
    Database Fundamentals
    The fundamental problems a database engineer thinks about in the shower
    21 1
    Nov 5, 2024
  18. 18

    In earlier eras of computing unreliable hardware forced system design discipline (checksums, retries, modularity).

    There is a lot of that we are rediscovering in the "Agentic Engineering" era!

    Wrote a small post about this idea.

    davidgasquez.com/reliable-unr...

    Agentic engineering is mostly about building reliable systems around unreliable components (like your friendly coding agent). A good analogy I like is how early computers were powerful, but not trustworthy enough to be used “raw”. Hardware failed, bits flipped, and storage was noisy. Engineers had to build systems around the machine to make things more predictable. And, they came up with a bunch of interesting ideas! Error correction codes Redundancy Checksums Validation layers Retry logic With these, even though reliability was not a property of the machine/computer, it became a property of the system. This is something close to where we are with “Agentic Engineering”. As you’ve probably experienced, a coding agent can be fluent, useful, and very wrong at the same time! The key is to, like the early programmers did, treat agents as noisy components instead of trying to get the perfect/bigger one.
    20 7
    Mar 24, 2026
  19. 19

    Wanted to write a Bluesky Python client this weekend. Realized there are a bunch of them much better written than I could! Sharing some of them.

    - https://github.com/jbn/psychonaut from @generativist.xyz
    - https://github.com/kiliczsh/gokyuzu from @kilic.bsky.social

    20 7
    Apr 24, 2023
  20. 20

    Figured out how to run @duckdb.org queries in Astro MDX! Query is executed at build time and data rendered into a plain HTML file (no JS).

    Very useful to reference remote datasets (e.g: doing dashboards) while keeping things minimal.

    davidgasquez.com/duckdb-in-as...

    Rudimentary, but it works!

    Table with data about random orders Astro DuckDB component
    19 2
    Jan 3, 2025
  21. 21

    Thanks to uv speeds, I've been writing many small scripts and abusing uv to run them instead of having them installed on my machine.

    This one can get metadata and data from Spain INE.

    github.com/davidgasquez...

    Not the most secure thing but very frictionless!

    19 0
    Nov 8, 2024
  22. 22

    Very aligned with the classic "Remove the legend to become one".

    www.eugenewei.com/blog/2017/11...

    Remove the legend to become one — Remains of the Day
    When I started my first job at Amazon.com, as the first analyst in the strategic planning department, I inherited the work of producing the Analytics Package. I capitalize the term because it was both...
    David Gasquez @davidgasquez.com
    Awesome resource on plotting (with Python's Matplotlib) small multiples with highlighted lines and annotations! 🤩 python-graph-gallery.com/web-small-mu...
    Small multiple with highlighted lines and annotations
    This post explains how to build a custom scatterplot with varying color, transparency, and size in Matplotlib to explore the relationship between the characteristics of astronauts and space missions. ...
    19 1
    Nov 3, 2024
  23. 23

    Wrote a post after spending one week using @zed.dev.

    davidgasquez.com/trying-zed-e...

    TLDR: Would love to switch once they work on a few extra features (Notebooks, Agent Mode, ...)

    David Gasquez @davidgasquez.com
    So tempted to try @zed.dev again... If this works great, the only thing I would miss is the Jupyter Notebook support. Perhaps that's a sign to also move to @marimo.io though! 🤷
    17 2
    Feb 21, 2025
  24. 24

    Excited about more people knowing about tools / ideas I've been working with in the last few years.

    - Content addressed storage
    - Fun data structures
    - Non blockchain decentralized networks
    - Local first approaches
    - Open protocols towards common digital infrastructure

    Simon Späti 🏔️ @ssp.sh
    Amazing how we get back to the open web with bsky. Great talk. Love the concepts. Anybody knows when the SQLite dbs come into play, in the intermediate aggregator, or on our local client most probably?
    17 3
    Oct 30, 2024
  25. 25

    Great showcase of the kind of work that usually goes unnoticed in the data world.

    Improving data quality is a hard collaborative effort. I am impressed by the standards @ourworldindata.org sets given the large amount of different datasets and sources they wrangle with!

    Our World in Data @ourworldindata.org
    Spotting and fixing data issues: how we help improve data quality on and off our publication
    Line chart showing an example of a data error that was detected and flagged for correction. The old data has a large spike in the middle of the timeline, while the new data shows a consistent line without the anomaly.
    16 2
    Nov 27, 2024