Eduardo Romero @foxteck

Let me know if this is just me: Noticed someone I know who is very "AI-pilled" and uses agents 24/7 to... start to talk noticeably more like these LLMs write. Eg more heavily using adjectives like "geniune", frequently terms like "the shape of" and many more examples

Whoaaa @max__drake showing a DAG of tasks with dependencies, where each task can be run by a coding agent fairy Perhaps the first spatial orchestration thing I’ve seen where space is used meaningfully

We're hiring! If you want to build tools developers reach for, come build with us. We've got multiple roles across engineering and are investing in Pydantic Logfire, the tool teams use to trace, evaluate, and improve AI apps and agents, from first prototype through production and everything after. → Agent Infrastructure Engineer (AI agents and sandboxes) → Evals & Continuous Learning Engineer → Observability Experience Frontend Engineer → Principal DataFusion Engineer (data store) → Principal Software Engineer (observability ingest) → Technical Program Manager Why join? Your ideas set the agenda instead of getting buried in process. You'll work with some of the best Python and full-stack developers around. And we're remote-first, getting together in person about once a quarter (usually abroad). Explore open roles and apply: pydantic.io/kyVbU

Postgres is the lowest common denominator for databases, and, to be honest, that is why it works so well as a default. If you do not need anything specific, just use Postgres. It is a super simple choice that just works. Personally, MySQL is still my first preference, but Postgres is good enough for almost everything a typical application needs. The pattern breaks at scale - with "scale" being super subjective here. Once your system grows beyond a certain point, you will have to opt for something more specific and purpose-built for the exact problem you have. A time series workload wants a time series database. A search-heavy workload wants a search engine. A graph-heavy workload wants a graph database. Vector workloads need a vector database. But, if you are starting, like we have always done, start generic and then go specific when the workload demands it. Hope this helps.

@HoytEmerson This is our current data stack and works great. We query with DuckDB from S3 (and pg occasionally) then do "in app" operations with polars. Transferring between DuckDB and Polars happens via Arrow (transparent to us) and we've been thinking about using Arrow with consumers too.

This is a great read: dbreunig.com/2026/05/04/10-… 10 Lesson for Agentic Coding

Had a blast with this bright, inquisitive group of current and former professional athletes yesterday. What a pleasure to sit down with the @NextLegacy_ fellows to discuss angel investing and company building. A wonderful community. Thanks for having me @ryannece & team!

Tomorrow @sh_reya and I kick off this free AI product engineering mini-course. Topics covered over 12 talks: 1. Design/UX & Evals 2. Retrieval 3. When & how to use open models effectively With these legends: @TheZachMueller @bclavie @xeophon @GoAbiAryan @barrowjoseph @willccbb @marek_galovic @beirmug Sign up & learn more here maven.com/lls/21f487

we just shipped @𝚜𝚞𝚙𝚊𝚋𝚊𝚜𝚎-𝚕𝚊𝚋𝚜/𝚝𝚊𝚗𝚜𝚝𝚊𝚌𝚔-𝚍𝚋 ◆ live queries ◆ optimistic mutations ◆ automatic Realtime sync ◆ fully typed shout out to @tannerlinsley - a very impressive framework. Currently in alpha:

𝗣𝘆𝗱𝗮𝗻𝘁𝗶𝗰 𝗔𝗜 𝘃𝟮 𝗶𝘀 𝗵𝗲𝗿𝗲, and your agents have never been more capable. The inner loop of an agent is settled, the leverage is in the layer around it, and v2 turns that whole layer into one thing you compose: the capability. It bundles instructions, tools, hooks, and model settings into a single unit. Plus a leaner core and the Pydantic AI Harness.

Really excited to see chDB become a named launch partner for AWS Lambda MicroVMs. When we started chDB, the idea was simple: what if every app, script, sandbox, and agent session could have its own private ClickHouse engine? Lambda MicroVMs makes that model feel very natural. Private compute. Private SQL engine. Fast start. Suspend/resume. No shared server in the middle. This is exactly the kind of infrastructure AI agents will need.

chDB @chdb_io

2 weeks ago

Excited to share that chDB is a named launch partner for AWS Lambda MicroVMs. chDB brings an in-process ClickHouse engine to every isolated workload — installed with a single pip install chdb. Together with Lambda MicroVMs, each workload can have its own private SQL engine with

0 5 24 13K 13

Pydantic AI version 2.0.0 is out! 🎉 github.com/pydantic/pydan…

Standard Metrics just wrapped an incredible all-company offsite in Denver We also closed 3 new hires over the past week in engineering, sales, & data solutions Passionate about accelerating innovation in the private markets? Apply to @metrics_co! standardmetrics.io/careers/

Yesterday we launched Automations in Kiro 🚀 Recurring agentic tasks, connected to your codebase Some fun ideas: • Obsidian vault gardener - links orphan notes, surfaces forgotten ideas • Arxiv paper monitor - detect new papers matching your research interests • Docs drift fixer - catches and updates stale docs • Morning briefer - overnight PRs, issues, and CI failures in one digest • Changelog generator - summarize recent changes • Flaky test fixer - analyze recent CI runs

There are so many cool things coming for Polars and Polars Distributed. The context switches to follow them all during the day are getting quite demanding. 🙈 All actively being worked on: # Polars - Polars 2.0; This release will make the streaming engine the default and fix some issues that block our development. For most users this transition will be smooth. - Out Of Core(Finally!); You will be able to set a memory limit and above that limit, Polars will start spilling morsels to disk. First PR: github.com/pola-rs/polars… - Drastically increasing our s3 throughput; With a fast engine, loading data from s3, bandwidth is the bottleneck and our top priority is saturating that fast with a adaptive algorithm that keeps tuning. # Polars Distributed - Distributed Expressions; This one deserves a blog post, maintaining information on ordering, partitioning, length and cost all at the same time has almost driven some team members to madness. - Distributed iceberg sinks and atomic writes; With roundtripping iceberg, we can progressively run on snowflake and bigquery data. To keep track of the 2.0 release, here is our milestone: github.com/pola-rs/polars…

Technology Radar Vol. 34 is live 👉 tinyurl.com/3ysepx95

MariaDB now has a DuckDB storage engine 🦆 `ENGINE=DuckDB` stores data in columnar format, pushes full queries to DuckDB's vectorized engine in-process — no ETL or separate clusters required: mariadb.org/duckdb-storage…

Today, HUD is excited to share our Series A funding! We are the platform for building high quality post training datasets. Over 50 businesses use HUD to build RL environments, sell them to AI labs, or train their own models from them. Our mission is to enable a generation of data entrepreneurs. The previous generation built apps to impact the world. We believe people will build infrastructure around data, both digital and physical, that align AIs to their specific goals. We’ve raised $16M in total funding from @Standard_Cap, @ycombinator, @ExceptionalCap, @Liquid2V, @twentytwovc , amazing angels such as @dylan522p from @SemiAnalysis_, @tszzl, @ivanburazin, @theo, and more!

Super excited to see the Beyond Benchmarks 2026 report from Emergence Capital go live, featuring Standard Metrics benchmarking data! @metrics_co 🤝 @emergencecap A favorite finding of mine: the growth vs. efficiency tradeoff disappears at the top. Top decile companies are growing 2–4x faster than top quartile at every stage, and the ones growing fastest are also the most capital-efficient. @kochiko2001 on our team chatted with Emergence GP, @thejoefloyd, about why the top decile is the place to watch in the video below. Give it a watch and check out the full report here: emcap.com/beyond-benchma…

Step Functions can now run Bedrock agents as a native workflow step! No more Lambda-wrapping-an-agent 🫰 You can drop agent reasoning directly into a state machine, right next to your Lambda, ECS, DynamoDB, whatever task! Cool things