Exa

Search + API

Exa is the discovery layer I hit when I need fresh research, engineering writeups, or niche demos that normal search misses. It’s a semantic search engine and API that crawls papers, blog posts, X threads, GitHub repos, videos, and more, returning high-signal results you can pipeline into agents or dashboards. (exa.ai)

Why it helps

• Research-grade recall. Queries surface up-to-date papers, code, and product writeups rather than SEO cruft. Perfect for staying ahead on AI/agent techniques.
• Structured metadata. Each result comes with summary text, categories, source URLs, and sometimes extracted entities, so it drops neatly into Notion, knowledge bases, or agent contexts.
• API friendly. You can hit the Exa API from your own tools—gathering context for video scripts, newsletters, or agent workflows.
• Multi-modal coverage. It indexes long-form posts, GitHub repos, Product Hunt launches, X threads, and recording transcripts, so you can triangulate what’s happening fast.

How I use it

1. Content research. Feed Exa queries into my video prep to pull the latest examples, benchmarks, or opinions.
2. Agent knowledge. Pipe Exa search results into agents so they have fresh references instead of static doc dumps.
3. Competitive scans. Monitor emerging AI products or frameworks and collect the best explainer content automatically.

Getting started

• Create an account at exa.ai and grab an API key.
• Use their hosted search UI for quick queries, or call the REST API to integrate results into your own apps/agents.
• Combine with embedding systems or note-taking tools to build living research hubs.
• Keep an eye on rate limits and data freshness—Exa ships frequent index updates so re-run queries as topics evolve.