TJ
@TJkrusinski
Followers
651
Following
10K
Media
106
Statuses
312
content at @trychroma
Menlo Park, CA
Joined April 2009
could it be that easy?
0
1
3
Sora 2 correctly diagnosing medical imagery when you supply it as the starting frame and the prompt of “diagnose this”.
1
2
6
Made some ad creative this week for @trychroma. @midjourney for stills and video @elevenlabsio for voice over and sound effects. Edited together in Davinci.
0
0
2
Agents perform better with ground truth, so we built a tool to help them get that.
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
0
0
2
The company I work at has this neat property where if a customer gives us one dollar it costs less than that one dollar to provide our service to them.
4
1
21
Using @bunjavascript makes me excited to build again, so easy, has everything and just works
6
0
51
I assure you more data was embedded today than yesterday, and tomorrow more yet than today.
0
0
3
I worked at Meta for 8 years, the consistent narrative among the noisy tech crowd on twitter was that FB was dying, yet every quarter Meta grew in revenue and DAU across FB and IG. I now work at a company that makes a vector database. Everyday on twitter people say RAG is dead.
1
0
7
You can't just yeet everything in the context window. Check out our research report on the topic.
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
0
1
5
hosting an event in NYC on July 7 on context engineering thanks to @tryramp for hosting https://t.co/Go2YVcVCm4
luma.com
Context engineering Context Engineering: "the art of providing all the context for the task to be plausibly solvable by the LLM." - @tobi Learn how leading NYC…
4
7
74
We introduce representative generative benchmarking—custom eval sets built from your own data that reflect real user queries. thank you @weights_biases for collaborating! link to report in replies
12
22
288