21. B Better prompts won't fix a broken context pipeline. Reliable agentic RAG depends on controlling ... Better prompts won't fix a broken context pipeline. Reliable agentic RAG depends on controlling 𝘄𝗵𝗮𝘁 𝗶𝗻𝗳𝗼ᵟ 08/21 00:02 ♡
22. Most search agents wait for instructions. Ours now reads the room first. The latest update is a bi... Most search agents wait for instructions. Ours now reads the room first. The latest update is a big one: 𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗱 &# 08/19 00:01 ♡
23. Why do creative workflows become harder to search as projects grow? Folders begin with a clear stru... Why do creative workflows become harder to search as projects grow? Folders begin with a clear structure. Tags work when teams have time to maintain them. Keyword search works when 08/13 23:01 ♡
24. Query Agent Search Mode now makes the recall-versus-precision tradeoff configurable. Search Mode re... Query Agent Search Mode now makes the recall-versus-precision tradeoff configurable. Search Mode rewrites a natural-language request into one or multiple Weaviate queries, each con 08/12 22:01 ♡
25. Scaling Test-Time Compute is the easiest way to get higher quality outputs from AI systems, but it c... Scaling Test-Time Compute is the easiest way to get higher quality outputs from AI systems, but it comes at a cost! ⚖️ How much extra compute you want to trade for quality depends 08/11 23:19 ♡
26. Your database can now speak MCP. Compatible clients such as Claude Code, Cursor, and VS Code can co... Your database can now speak MCP. Compatible clients such as Claude Code, Cursor, and VS Code can connect directly to `/v1/mcp` on the same port as the REST API, without operating a 08/06 21:36 ♡
27. Most AI chat interfaces fail before the first message is even sent. Here's the small but powerful f... Most AI chat interfaces fail before the first message is even sent. Here's the small but powerful fix we just added to the Weaviate Query Agent. The Query Agent is your AI-powe 08/05 18:07 ♡
28. HTTP 200 isn't where you stop validating an import. A batch can succeed while individual objects fa... HTTP 200 isn't where you stop validating an import. A batch can succeed while individual objects fail due to rate limits, vectorizer errors, or schema mismatches. Those issues 08/04 21:20 ♡
29. How much creative work gets recreated simply because nobody can find the original? A designer remem... How much creative work gets recreated simply because nobody can find the original? A designer remembers the mood but not the filename. An editor knows the shot exists but not which 07/30 23:15 ♡
30. Query Agent just got 5-10% better at recall@5. Same cost. Same speed. Nothing you did. Query Agent... Query Agent just got 5-10% better at recall @5 . Same cost. Same speed. Nothing you did. Query Agent now uses OpenAI's new 𝗚𝗣𝗧-𝟱.𝟲  07/29 23:00 ♡
