The role, as published
About the company
Our client is a fast-growing software company.
The role
Raydar is recruiting for this role on behalf of our client. Take ownership of core infrastructure that helps an automated software agent run faster and more dependably. This is a hands-on individual-contributor position that blends systems engineering, experimentation and direct conversations with end users, in close partnership with company leadership.
What you'll do
- Own and evolve the systems that let the agent retain and manage context over time.
- Build and refine an automation framework that drives web browsers, with emphasis on navigation and dependable interaction.
- Work through obstacles that arise when automated traffic is screened by websites.
- Cut token consumption by swapping long natural-language instructions for deterministic, rule-based logic.
- Create test environments and performance benchmarks that surface regressions and track how well the agent handles realistic tasks.
- Talk with users to learn where the agent struggles and turn their feedback into shipped fixes.
- Refine prompts so the agent behaves consistently and efficiently in varied situations.
What we're looking for
- At least 1 year of full-time professional engineering experience; internships do not count toward this.
- A track record building stateful memory components, browser automation tooling or similar infrastructure for AI agents.
- Comfort editing prompts at a detailed level and explaining the reasoning behind each adjustment.
- Based in San Francisco or willing to relocate, working in the office five days per week.
Bonus points
- Earlier experience as a founding engineer in an unstructured startup setting.
- Ongoing open-source contributions, ideally to a project with a sizable following.
- Experience building or improving systems powered by large language models.
Location and work model
- San Francisco, CA, United States
- On-site, 5 days per week in office
- Full-time
Compensation, as the posting describes it
- Equity