Join @jeffreyhuber and @dbreunig in a wide ranging conversation about context engineering and agents. https://t.co/zhi3lSuRu2

Watch the video on YouTube. https://t.co/RHQGOhkvzE

Using open source Chroma in production? We want to send you some Chroma swag. https://t.co/iaAUxGIXR9

Today, we’re launching: - Support for sparse vectors including BM25 and SPLADE - A new Search() API with support for hybrid search, including reciprocal rank fusion or your own custom weighting

Introducing Chroma Package Search Enable your AI agents to search the source code of your package dependencies Includes 3 MCP tools: package_search_hybrid() package_search_grep() package_search_read_file() Add it to Cursor, Claude Code, VS Code, and Codex in 5 seconds

Link → https://t.co/b1el8XC9OO

A deep dive on wal3 - Chroma's open-source write-ahead log built on object storage. Featuring: - a 30-year-old lock-free algorithm - Amazon S3's newest conditional writes feature - a novel checksumming technique called setsum

Every successful AI company you know is great at one thing 👉 context engineering. Great talk from @jeffreyhuber on the @latentspacepod who covers a wide range of context engineering topics: - context curation - retrieval, chunking, indexing - re-rankers - evals - memory

Introducing Chroma Cloud: an open-source serverless search database that is fast, cost-effective, scalable, and reliable.

Collection forking is now available on Chroma Cloud! Call fork() on a collection to instantly get access to a fork, accruing storage costs only on new data, enabled by Chroma's object store backed storage engine. Use it for dataset versioning, checkpointing, or syncing coding

Chroma Collections now support regex search. No extra tooling or configuration to set. Regex optimized indexes prevent full scans in most scenarios. Available now on Chroma Cloud, link below.

Go behind the research with @kellyhongsn and @atroyn. Our latest technical report: "Context Rot" investigates how model performance grows increasingly unreliable as input length grows.

Introducing our latest technical report: Context Rot - How Increasing Input Tokens Impacts LLM Performance Our results reveal that models do not use their context uniformly. full report in replies

We can’t call it “engineering” if we can’t predict when it breaks. Context engineering turns vibes-based prompting into systems thinking. Chroma CTO, Hammad Bashir shares his thoughts on Context Engineering.