About the role
Sorce is building AI agents that apply to jobs for users. Users swipe through jobs like Tinder, and our AI applies for them automatically. The core of the product is a browser agent that fills out and submits job applications across hundreds of different ATS systems. We're looking for an engineer to own the reliability of that agent end to end.
This is a role for someone who likes fixing hard problems. The long tail of ATS systems is a mess. Applications fail for all kinds of reasons. Your job is to understand why, fix it, and make sure it doesn't happen again.
You'll own the agent and everything that flows through it. Submission rate is one of our most important product metrics, and you'll be the person moving it.
What you'll do
- Live inside the agent code: diagnose why applications fail, fix the failure modes, prevent recurrence
- Triage the hold / expired / failed buckets every day. Build the failure taxonomy, automate retries where possible
- Improve form-fill reliability, captcha handling, and edge-case behavior across the long tail of ATS systems
You'll be a good fit if you:
- Ship fast and operate with urgency. We’re lean and we move fast.
- Have experience with llm-driven browser agents and browser automation (Playwright, Selenium, Puppeteer, Stagehand, Browser Use, etc).
- Have experience with browsers: anti-botting, captchas, proxies, CDP, caching, cookies, etc
- Are strong in Node.js, TypeScript, and Python.
- Are comfortable in Postgres. You'll be querying application tables constantly.
Bonus if
- You have familiarity with Golang
- You’ve dabbled with Captcha-solving: commercial solvers, custom approaches, vision-model-based
