Senior Software Engineer (Formal Methods & Agentic Systems)
New
A
AZXArtificial Intelligence
RemoteFull-TimeSenior
Salary$140K - $230K; $140K – $230K • Offers Equity • Offers Bonus
Apply NowOpens the employer's application page
Job Details
- Experience
- 5+ years
- Required Skills
- PythonFastAPIPostgres
Requirements
- 5+ years of productionization experience: taking prototype or research code to production with packaging, tests, CI, observability, and documentation.
- Real depth in formal methods (SMT/constraint solvers like Z3, automated theorem proving, or heuristic search).
- Experience with data modeling and shape validation (ontology/taxonomy design, SHACL, RDF/OWL/SPARQL, or knowledge graph schemas).
- Agentic AI literacy: understanding LLM agent failure modes and agent-solver boundaries.
- Strong generalist engineering skills with Python fluency and service design.
- Practical familiarity with core stack: Python 3.12+, FastAPI, and Postgres.
- Experience with test engineering for formal systems (counterexample regression testing, property-based testing).
- Working knowledge of LLM provider APIs and agent frameworks.
- Bachelor's degree (Master's is a plus).
- Ability to travel 2x/year for company summits.
Responsibilities
- Take the research-grade formal/agentic system to production: a real package, test suite, service interface, documentation a cold-joiner can use, and a release cadence.
- Own the architecture for reliability, packaging, test coverage, typing, CI, performance, and release discipline of the formal/agentic system.
- Wrap solver runs in agent loops where the agent proposes and the solver disposes, deliberately defining what the agent may touch when a proof fails.
- Model messy client business rules — compliance requirements, rate structures, program eligibility, design constraints — as constraints and shapes that check mechanically, and build the review habit that keeps those models honest.
- Reason over per-customer digital twins, checking proposed changes against the twin's constraints and shapes before anyone touches the real system.
- Define the agent seam: where LLM agents may assist (translation, hypothesis, explanation) and where they're forbidden (anything that asserts).
- Make proof results legible to client stakeholders who will never read a proof — clearly communicating what was checked, against what, and what was not checked.
View Full Description & ApplyYou'll be redirected to the employer's site
