Elvis S. — Founder at DAIR.AI | Investor | Prev: Meta AI, Galactica LLM, Elastic, Ph.D. | Serving 7M+ learners around the world
Founder at DAIR.AI | Investor | Prev: Meta AI, Galactica LLM, Elastic, Ph.D. | Serving 7M+ learners around the world
Elvis S. ranks #63 of 18,566 LinkedIn creators in Research, and is a standout voice in Belize. They have 88.0K followers and published 46 posts in the last 60 days at a 0.3% average engagement rate.
- 88.0K followers
- 46 posts / 60d
- 0.3% avg engagement
- 1.6K follower growth / 30d
The roast
Elvis democratizes AI research the same way a guy in a trench coat outside a middle school democratizes fireworks. He claims to serve seven million learners, yet with an engagement rate under 0.3%, he is clearly the only student who didn't get the memo that the class was cancelled.
About Elvis
Building DAIR.AI, wherein we are democratizing AI research, education, and technologies. Our mission is to enable the next generation of AI innovators and creators. I am an AI researcher focusing on training LLMs, information retrieval, AI Agents, RAG, and related fields. I provide consulting services and advise some of the top AI companies/startups around LLM-powered applications, AI agents, AI research, go-to-market, insights, product, and growth strategies. I previously worked on the Galactica LLM, Papers with Code, PyTorch, Meta AI, and Elastic. Connect with me on Twitter: https://twitter.com/omarsar0
Highlights
- Top 1% Audience — 87,983 followers
- Top 1% Creator — 46 posts in 30 days
- Top 5% in Research — Ranked #1 of 57 creators
- Top 5% Impact — 279 avg engagements per post
Recent posts
// Persistent Workspaces for Long-Lived Claude Code Agent Teams // Four issues to be aware of: > Working state vanishes when a terminal closes and the team cannot be resumed. > Compaction condenses the conversation and blurs each agent's working detail. > Decisions and operations get trapped in compacted chats, which accumulates as agentic technical debt. > Every handoff needs another long prompt. ATWZ is a filesystem operations layer built around Claude Code's native Agent Teams. Each agent gets a workstation directory holding its working state alongside the skills, hooks and scripts
25 reactions · 4 comments · 0 reposts
Neat work on long-horizon agents. Splitting a hard task across agents is typically how standard multi-agent work. The usual design lets them exchange findings only at phase boundaries, through staged handoffs or synchronized rounds, so communication and work stay mutually exclusive. AgentRadio breaks that constraint with an asynchronous message-passing layer. Three primitives, threads, messages, and waiting for mentions. The last one runs as a background task, surfacing a teammate's discovery without interrupting foreground work. On SWE-Atlas QnA, a single Claude Code agent on Opus 4.6 reso
79 reactions · 5 comments · 0 reposts
New research from Meta. (bookmark it) It's on how to fix agents that forget previously made decisions. It's well know that long-horizon agents keep forgetting decisions they already made. Meta researchers give this failure a name, behavioral state decay, where task facts, prior attempts, and open subgoals get buried in the context window or pushed past it, so they stop influencing the next action. Their fix runs a separate memory agent alongside an unmodified action agent. It maintains a structured memory bank from the recent trajectory and decides, each step, whether to inject a memory-gr
237 reactions · 6 comments · 24 reposts
Last updated 2026-08-01