About the Role
We are building the agent runtime that finds, validates, and prepares fixes for vulnerabilities automatically, while people retain judgment and merge control. This involves solving real distributed-systems and reasoning problems.
Responsibilities
- Build and harden Nullify's multi-provider agent runtime, including loops, structured output, budget governance, and checkpoint/resume for multi-hour runs.
- Design agents that reason about business logic and real application behavior, moving beyond simple pattern matching.
- Build and grow the evaluation harness, encompassing labelled test cases, scoring, and regression tracking.
- Own agents end-to-end, covering detection, validation, fix generation, and intermediate judgment calls.
Requirements
- Strong software engineering fundamentals in Go or Python.
- Experience with production systems, not just notebooks.
- Experience building or operating LLM-based agents beyond single prompt-response calls, including tool use, multi-step reasoning, retries, and evaluations.
- A scientific approach to measuring effectiveness before deployment.
- Genuine interest in security and a desire to become an expert.
Skills
- Go
- Python
- LLM-based agents
- Tool use
- Multi-step reasoning
- Retries
- Evaluations
- Security
