500 Skills, Zero Fine-Tuning: LinkedIn's Playbook for AI Agents That Actually Know Your Codebase
Ajay Prakash
- When
- Wednesday, July 111:40 AM – 12:00 PM · 20 min
- Where
- Track 8San Francisco, CA · imported from ai.engineer's public schedule feed
About this session
Everyone's building custom AI agents. We didn't. Instead, we built CAPTAIN — an MCP server that makes any off-the-shelf coding agent understand LinkedIn's entire engineering stack. The secret: a meta-tool architecture (discover → inspect → execute) and composable skills that encode tribal knowledge as executable workflows. 500+ skills later, it's used across all of LinkedIn engineering. I'll show you the architecture in 10 minutes and why context engineering beats model engineering every time.
Speaker
Senior Staff Software Engineer, Linkedin
Ajay is a software engineer at LinkedIn with 14 years of experience in software, building large-scale systems and AI. For the past four years, his work has shifted fully into AI: LLMs, prompt engineering, context engineering, and AI agents. He previously led AI platform and product initiatives for LinkedIn Sales Navigator. Over the past year, he's led efforts to improve the effectiveness of coding agents by connecting them to LinkedIn's internal tools and context, making them genuinely useful inside a large engineering organization. He now leads AI agent platform efforts at LinkedIn, the most interesting work of his career so far.
More in Context Engineering
- Build-Time vs. Run-Time: Why Your Dev Tools Will Fail in ProductionWednesday, July 1 · 10:45 AM – 11:05 AM · Track 8
- It’s Tokens All The Way Down: How RLMs are DifferentWednesday, July 1 · 11:10 AM – 11:30 AM · Track 8
- Your agents lack context: Here's how to fix "You're absolutely right!"Wednesday, July 1 · 12:05 PM – 12:25 PM · Track 8
- Prompt, Memory, Weights: The Architecture Decisions Most AI Teams Make by AccidentWednesday, July 1 · 12:05 PM – 12:25 PM · Expo Stage 4 SE
- How long can your skills be before your agent forgets what you told it?Wednesday, July 1 · 1:30 PM – 1:50 PM · Track 8
For developers: this programme is open data — JSON, iCal, schedule XML and an MCP endpoint.Show endpointsHide
- JSONEvery published session and speaker, in one request./aie-worldsfair-2026-import/feed.json
- iCalSubscribe in Google, Apple or Outlook Calendar./aie-worldsfair-2026-import/feed.ics
- Schedule XMLfrab / pentabarf — the format conference apps import./aie-worldsfair-2026-import/feed.xml
- MCP + RESTPoint Claude at the programme. OpenAPI 3.1 included./agents
No key, no signup, CORS open. Everything here is generated from the same data the organisers edit.