Thomas van Dongen

@thomas_v_dongen

Co-founder @minishlab. Working on Semble, Model2Vec, SemHash, Pyversity, and static embedding models.

github.com/pringled

Joined March 2011

119 Following

43 Followers

Replies Reposts Media

Thomas van Dongen
@thomas_v_dongen
Apr 28
We're releasing a new package today: Semble. It's a Local code search for agents with near-transformer accuracy. No API keys, runs fully on CPU, and it indexes repo's in ~250ms.
Minish
@minishlab
Apr 28
Today we're releasing Semble, a fast and accurate code search library built for agents 🤖! We're also releasing potion-code-16M, a small code-specialized static embedding model that powers Semble. 🧵
3
Thomas van Dongen
@thomas_v_dongen
Mar 10
I've just released a new package: agentcheck 🔐 When you run an AI agent in your shell, it inherits everything in your environment: cloud credentials, API keys, Kubernetes contexts, SSH keys. Agentcheck scans this efficiently so you can run an agent safely. 🧵
1
Thomas van Dongen
@thomas_v_dongen
Feb 2
I've just released Pyversity 0.2.0 📦 The biggest addition is a benchmarking suite for diversification rerankers (MMR, MSD, DPP, SSD) across 4 recommendation datasets. 🧵
1
Thomas van Dongen
@thomas_v_dongen
Jan 19
I've just released SemHash 0.4.0 🧹 The biggest new feature is multimodal support: you can now use SemHash to deduplicate and filter datasets of any modality (text, image, audio, and more). 🧵
1
Thomas van Dongen
@thomas_v_dongen
Dec 8, 2025
Last week I had the pleasure of joining the @weaviate_io podcast to talk about Pyversity. If you're interested in a deep-dive into diversification for retrieval, this might be worth a listen. Thanks for having me @CShorten30 !
Connor Shorten
@CShorten30
Dec 8, 2025
I am SUPER EXCITED to publish the 132nd episode of the Weaviate Podcast with Thomas van Dongen (@thomas_v_dongen), head of AI engineering at Springer Nature! 🎉 Thomas is the creator of Pyversity, a fast, lightweight open-source python library for diversifying retrieval results!
3