debabrata pattnayak —
hey @user what are you thoughts on implementing semantic cashing in archestra ?
#general
Hey, I've submitted a PR for issue #3218 — auto-sync permissions for Jira + Confluence. Would love any early feedback before review.
Hi everyone! I'm Harry, just joined the community. I've been looking at issue #4429 (automatic Pod restart on MCP server update). I've already done some local analysis and identified that
McpServerModel.update is missing the trigger to notify the runtime manager. I've also verified a fix with unit tests in mcp-server.test.ts. I'd love to take this on officially—is it okay if I /attempt it on GitHub?
I tried applying to some open issues in Archestra's repo.
It says only prior contributors can apply to issues.
Is it how it's supposed to be?
hey @user what are you thoughts on implementing semantic cashing in archestra ?
Read-only live mirror of Archestra.AI Slack
👋Join the discussion withAI enthusiasts!hey @user what are you thoughts on implementing semantic cashing in archestra ?
3 replies
Hi! Could you please add more details to your proposal?
(semantic cashing) It helps reduce latency and reduce cost by serving cached responce. We can implement vector similarity search to cache AI responce by serving cached results for semantically similar requests even when the exact wording differs
i will write detail proposal with implementation plan shortley.