Founding Member: AI Engineer

AI Security Assurance is an early-stage, remote-first startup on a mission to accelerate the delivery of trustworthy systems — systems proven secure and safe by design. We build automated tooling for whole-system security and safety assurance: it analyzes a system as it's actually built — including the interactions between components, where the most damaging design-level flaws hide — and produces the evidence that trustworthy systems require. Our work is grounded in Systems-Theoretic Process Analysis (STPA) and formal methods for security, following the NIST systems security engineering framework (SP 800-160). We're bootstrapped and pre-funding, building toward funded government R&D (DARPA, SBIR) and a commercial platform. We're assembling a founding team — people who want skin in the game.

About the Role

This is a founding-member role, not a salaried job today. We're looking for someone to help us build the company and win the funded research contracts that pay for the work. In place of a salary now, founding members share in the upside through equity / ownership participation; when we land a contract that funds the role, that contract compensates the work. We're honest about the risk — there is no guarantee of funding — and we're looking for people who want to build something real and share in its success.

As a founding AI Engineer, you'll build the AI layer of our tooling and the human-AI collaborative environment around it: multi-agent workflows, retrieval over system models and code, and the production software that makes all of it usable by real analysts. Our approach is deliberately neuro-symbolic. We use deterministic methods wherever possible and apply AI narrowly, where it earns its place, and never leave it in control of the result. Assurance only means something if the tool producing it can itself be trusted. You'll help define where AI belongs in a trustworthy assurance pipeline, build the evaluation that proves it works, and help shape the research behind our proposals.

Responsibilities

• Design and build multi-agent workflows and retrieval-augmented (RAG) pipelines over system models, documentation, and source code.
• Build the human-AI collaborative environment: an interface where analysts review, refine, and override what the system produces, with humans in control.
• Establish rigorous evaluation: build datasets, define metrics, and measure agent and model performance rather than assuming it.
• Work with our systems security and formal methods leads to decide which parts of the pipeline should be deterministic and which benefit from AI.
• Support both cloud and on-premises deployment, including secure and disconnected environments.

Requirements

• Hands-on AI / LLM engineering: multi-agent workflows, retrieval-augmented generation, prompt and context engineering
• Full-stack production software engineering (e.g., Python and TypeScript), not just notebooks and prototypes
• Experience building data-intensive, agent-driven applications and the interfaces around them
• Rigorous evaluation of model and agent performance (e.g., precision/recall, LLM-as-judge, building evaluation datasets)
• Sound judgment about where AI is and is not appropriate, especially where correctness matters
• Ability to work independently and wear multiple hats in an early-stage team
• Ability to work in the U.S. without sponsorship; willingness/ability to obtain a U.S. security clearance for future government work

Nice-to-haves

• Cybersecurity, safety-critical, or other high-assurance domain experience
• Knowledge graphs, code analysis, or program-analysis tooling
• Formal methods exposure (theorem proving, model checking)
• UX sense for expert tools, and experience with cloud and on-premises or air-gapped deployment
• Research, publications, or open-source work in AI/ML or security

This is a founding-member position, not a paid role at this time. AI Security Assurance is bootstrapped and actively pursuing funding. Founding members join now to help build the company and win funded research contracts, participating through equity/ownership rather than salary initially; paid compensation begins when a contract funds the role. There is no guarantee of funding.