What you’ll learn
Three tiers of alumni pools, founder-factory networks worth chasing warm intros for, communities, talent platforms, and a GitHub sourcing checklist.
Tier 1
verified FDE training grounds
Tier 2
high-yield via customer-facing engineering culture
Tier 3
heterogeneous quality, filter hard
Online communities:
Read these every week:
Where the best candidates live the worst on LinkedIn, mine contributor lists of LangChain, LlamaIndex, Inspect, Promptfoo, Braintrust, Arize Phoenix, Ragas, DSPy, Anthropic SDK, OpenAI Cookbook, vLLM, Modal, Baseten, MCP server repos.
GitHub sourcing checklist
Use to filter contributor profiles
Search Github for users with the following profile: PRs (last 90 days) merged into LangChain, LlamaIndex, vLLM, Modal, Braintrust, Arize Phoenix, Ragas, DSPy, MCP server repos. Look for: • Recent (last 90 days) commits to AI infra repos. • README quality on personal agent/RAG repos with architecture diagrams. • 'Compulsive shipper' pattern of many small repos vs. one bloated one. • Mixed languages (TypeScript + Python + Go) signaling full-stack.
Key takeaways