For recruiters and hiring managers hiring a AI Engineer (eval, guardrails, release side).
Role fit, with proof you can run
AI Engineer (eval, guardrails, release side)
Mid-level and senior IC roles. W2 full-time or contract-to-hire, Remote (US), based in Arizona.
AI features with the tests and release gates written first, so anyone can re-run the results.
Verify in sixty seconds
Each step below checks one claim behind the AI Engineer (eval, guardrails, release side) page.
- Open the release-gate case study and read what one failing assertion does to a merge
- Clone the public adversarial-gate repo and run:
python3 build_corpus.py && python3 evaluate.py
featurecontrol PASSplanted regression FAILgatereceipt
Requirements and proof
| Requirement | Proof |
|---|---|
| Build an applied AI system where every output is checked before it is trusted | Deterministic gate battery case study |
| Write the evaluation and release gate before the feature it protects | Release-gate case study |
| Orchestrate more than one agent and verify the combined result | Agentic fleet verification case study - Public attested-tool-results repo - Public agent-trace-receipts repo |
| Test an AI feature against adversarial input before it ships | Adversarial red-team suite - Public adversarial-gate repo - Public taint-flow-guard repo - Public RAG groundedness repo |
| Make a model-backed result reproducible so a re-run returns the same verdict | Artifact-lineage case study |
| Ship runnable code with a replay command | Public adversarial-gate repo |
Public repos that prove this title
- attested-tool-results
- agent-trace-receipts
- llm-adversarial-gate (claims p1, p2, p3, p6)
- taint-flow-guard
- rag-groundedness-gate
What could go wrong hiring me
Most proof here is self-built. None shipped inside a product team serving outside users.
Counter: Release-gate case study
The test corpus is self-written. No outside benchmark yet.
Agent work here is a one-person fleet. There is no multi-team platform.
Verify in 60 seconds
- Open the release-gate case study and read what one failing assertion does to a merge, /cases/eval-release-gate
- Clone the public adversarial-gate repo and run:
python3 build_corpus.py && python3 evaluate.py, https://github.com/mpuodziukas-labs/llm-adversarial-gate