AI Search Engineer
Build the search infrastructure that powers how AI agents find, understand, and reason over the open web. Engineered from the ground up for an agentic world.
Why this role matters
About the role
We are rebuilding search from the ground up for an agentic-first world. This isn't about wrapping an existing engine. It's about rethinking what search means when the user is an AI agent running a research workflow, not a human typing a query into a box.
This is one of the highest-responsibility and highest-leverage roles at Olostep. The opportunity here is to build the search infrastructure for the AI era, the equivalent of what Google built for the human web, but designed from day one for agents. The person in this role will have enormous impact on where the company goes.
You will own the full search pipeline: from how we crawl and index the web at scale, to how we process and rank results for both fast lookup and deep, multi-step research. You'll design systems that are reliable, low-latency, and built around the reasoning patterns of modern AI agents.
What you'll do
- Design and scale a distributed web crawler optimized for freshness, coverage, and extracting agent-relevant signals.
- Build and maintain a high-throughput indexing pipeline that can process billions of pages with structured metadata.
- Develop query processing and ranking systems that serve both fast search and deep research use cases.
- Build retrieval pipelines that are natively agentic, with awareness of context, multi-hop reasoning, and iterative query refinement.
- Instrument the search stack with observability tooling and use it to continuously improve relevance and reliability.