Senior Software Engineer (Formal Methods & Agentic Systems)
Company: AZX
Location: Seattle (Remote)
Salary: $140k - $230k per year
Type: Full-time
Remote: Yes
Posted: 2026-08-31
About this role
About AZX
Our mission is to accelerate positive impact in critical industries through AI transformation. We specialize in physics-informed ML and enterprise AI solutions that directly address climate and sustainability challenges.
We’re growing quickly and already work with category-leaders in real estate (CBRE), energy (LevelTen Energy), logistics (Flexe) and utilities.
We’re a public benefit corporation, founded in 2024, and have been profitable since inception.
We work on challenges in clean energy, decarbonization, climate risk, energy systems, and global economics. We’re building our company for long-term success and aim to create the ultimate place to work for those passionate about AI and making a positive impact.
About This Role
We are seeking a Software Engineer to help build and implement models, using agentic problem-solving wrapped around formal methods, that solve client issues that have strict compliance rules, tariffs, equipment constraints and much more . You'll partner directly with the system's inventor, learn the design deeply, and take it to production — packaging, testing, CI, documentation, and the interface that lets every AZX engineer put provably-right answers into client solutions. This is the formal solutioning layer of the stack: the part that produces checked answers when a client's problem actually has one. You'll sit between research and production, comfortable in both, and help shape where the system goes next.
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 ...