Why This Role Matters: 

We're building an agentic platform — an LLM agent runtime that reviews code, writes tests, and builds from tickets on real development repos — alongside the services that power it: a control plane that orchestrates agent runs, an intake registry that routes work requests through to automated fulfilment, and the integrations that tie them together. Someone has to make sure all of it actually works. This role owns that question across the stack — from "did the agent's code review make sense?" to "did the API return the right thing?" to "did the end-to-end flow complete without dropping state." You're the signal between what we ship and what's actually good. 

How You'll Contribute: 

• Test the services we build — control plane APIs, intake registry, agent dispatch flows, webhook integrations — at the API and integration level. 

• Review agent output — code review comments, generated tests, draft MRs — and assess accuracy, relevance, and quality against the codebase and context. 

• Write and maintain test cases and evaluation criteria: API contracts, end-to-end flows, agent output quality for each agent type (reviewer, tester, builder). 

• Identify failure patterns across the stack: broken API responses, dropped state, hallucinations, missed issues, bad diffs. Log them clearly. 

• Work with the platform team to translate findings into fixes, guideline improvements, and guardrail changes. 

• Own tickets end-to-end: test plan → findings → documented change → verified improvement. 

• Sit in on design reviews; help define what "done well" looks like before work ships. 

What We're Looking For: 

• 1-3 years of professional experience with software testing, QA tooling, or API test automation (Postman, pytest, httpx, or similar).
• Attention to detail and the ability to read code across at least one language (Python or TypeScript preferred). 

• Comfortable with Git, HTTP/REST APIs, basic SQL, and Linux shell. 

• Writes things down clearly — test cases, failure patterns, improvement proposals. 

• Asks questions and communicates what's wrong without waiting to be asked. 

​

Nice-to-haves (all optional):

• Familiarity with LLMs or AI agents
• Python or TypeScript · any CI/CD exposure
• AWS basics (ECS, CloudWatch, Secrets Manager)

What you'll gain:

Ground-floor experience on a production agentic platform and the services behind it, a direct line between your findings and what ships, a senior mentor, growth path to a QA or product engineering role, paid AWS cert if wanted.