Jaixii | Building Engraphis

@CodingDevTools

Building @Engraphis, local-first memory for coding agents. I write about agent context, open source, and developer tools.

Building in public

github.com/Coding-Dev-Too…

Joined March 2018

Build in Public
Jaixii | Building Engraphis
@CodingDevTools
52m
Decisions need more than a match. Before an agent changes code, can it inspect the source, scope, and whether it’s still current? Engraphis is local-first memory for AI agents: durable, scoped, explainable recall. New features soon engraphis.com #AI #Developer
Web Developers
Jaixii | Building Engraphis
@CodingDevTools
Aug 5
Give your agent more room to think. Daily updates. Engraphis benchmarks: 98% fewer tokens on long history, 73% less retrieved memory per Q, 55% less on full tool responses, 53x more evidence per token budget. engraphis.com #AI #Engraphis #LLM #Developer #Coding
Jaixii | Building Engraphis
@CodingDevTools
Jul 31
I've done so many updates on #Engraphis this week I haven't had time to post. It's really good and continues getting updates daily. The software helps reduce LLM token and context costs by Up to 98%. Any support is appreciated! engraphis.com #DevTools #AI #Coding
Machine Learning & AI
Jaixii | Building Engraphis
@CodingDevTools
Aug 5
Give your agent more room to think. Daily updates. Engraphis benchmarks: 98% fewer tokens on long history, 73% less retrieved memory per Q, 55% less on full tool responses, 53x more evidence per token budget. engraphis.com #AI #Engraphis #LLM #Developer #Coding
Jaixii | Building Engraphis
@CodingDevTools
Jul 31
I've done so many updates on #Engraphis this week I haven't had time to post. It's really good and continues getting updates daily. The software helps reduce LLM token and context costs by Up to 98%. Any support is appreciated! engraphis.com #DevTools #AI #Coding
Build in Public
Jaixii | Building Engraphis
@CodingDevTools
Aug 5
Give your agent more room to think. Daily updates. Engraphis benchmarks: 98% fewer tokens on long history, 73% less retrieved memory per Q, 55.38% less on full tool responses, 53x more evidence per token budget. engraphis.com #AI #Engraphis #LLM #Developer #Coding
Jaixii | Building Engraphis
@CodingDevTools
Jul 31
I've done so many updates on #Engraphis this week I haven't had time to post. It's really good and continues getting updates daily. The software helps reduce LLM token and context costs by Up to 98%. Any support is appreciated! engraphis.com #DevTools #AI #Coding
Build in Public
Jaixii | Building Engraphis
@CodingDevTools
Jul 31
I've done so many updates on #Engraphis this week I haven't had time to post. It's really good and continues getting updates daily. The software helps reduce LLM token and context costs by Up to 98%. Any support is appreciated! engraphis.com #DevTools #AI #Coding