Distinguished Engineer, R&D Lead

Join Keystone AI to create a safer digital landscape by assisting companies to secure their software through developer-led practices! Keystone AI ’s mission is to establish new standards for coding that transform the ways software is created. We do this by providing the world’s leading agile learning platform that delivers the most effective secure coding solution for developers to learn, apply, and retain software security principles. More than 600 enterprises trust Keystone AI to implement agile learning security programs, build safer software, and create a culture of developer-driven security. We are a venture-backed company with offices in Australia, the United States, Belgium, Iceland, and the United Kingdom. Software development is moving through its biggest shift ever — from human-written code, to AI-assisted development, to fully agentic systems where AI writes, tests, and deploys code autonomously. At every stage, enterprises face the same three problems: developers and agents need to learn to build securely, someone needs to govern what AI can and can’t touch in the codebase, and when something goes wrong, you need to know which AI did what, where, and for whom. Keystone AI is building the control plane for secure AI software development. We are defining the AI Software Governance category — and this role exists to ensure the technical foundations of that category are sound, forward-looking, and defensible. As our first Distinguished Engineer and founding R&D Lead, your mandate is broad and deliberate: tackle the hardest unsolved problems on our roadmap, generate the technical intelligence that keeps us ahead of the market, and build the prototypes that prove — or disprove — the ideas that will define our next generation of product. Key Responsibilities You own three things. How you balance your time across them will shift as the landscape evolves and as you develop your own view of where the highest-leverage opportunities lie. 1. Solve the Hard Problems The roadmap contains technical challenges that can’t be solved by delivery teams working in two-week sprints. These are ambiguous, cross-cutting problems that require structured discovery before anyone can commit to building them. You lead time-boxed R&D cycles to produce prototypes and clear recommendations: build, explore further, or stop. Current challenges in this space include areas like unified risk modelling, signals for MCP and other services, and AI code detection accuracy, but you will shape this backlog in collaboration with the rest of the teams, not merely execute it. 2. Push into the Whitespace This is genuinely forward-looking work. As developer workflows move toward fully agentic development, new technical questions are emerging that the industry hasn’t answered yet. How do you convert human security expertise into formats that AI coding agents consume natively? How do you govern an autonomous agent that writes, tests, and deploys code? What does code provenance mean when five agents contribute to a single commit? How should enterprise security intelligence integrate with the MCP ecosystem? SCW holds thousands of vulnerability patterns and secure coding challenges across dozens of languages — the raw material for a new class of agent-consumable security intelligence. You own the technical discovery of how to deliver that intelligence, working in close partnership with our content and security domain experts. The agent governance space is whitespace. The competitive window is narrow. Your job is to explore it faster and more rigorously than anyone else in the market. 3. Keep us Pointing Forwards The control plane roadmap is sequenced against a model of developer AI adoption, but the market moves fast enough that the model can become stale. You provide the structured intelligence that keeps the product organization pointed at where developers are heading, not where they were six months ago. This means tracking real-world developer workflows, monitoring competitive dynamics, scanning emerging technology and academic research, and synthesizing it into regular outputs that directly inform product and content strategy. You will publish a monthly Developer Workflow Pulse and a quarterly Landscape Brief — these become core inputs to roadmap decisions and, over time, raw material for SCW’s external thought leadership. How you Operate The Frontier Pod Model Your primary vehicle for structured discovery is the Frontier Pod — a temporary, cross-functional team formed around a single named problem for a six-week cycle. You convene the pod, define the problem brief, draw in the right expertise from across product domains, and present findings at the end of each cycle with a clear recommendation. This model gives you access to the right people for each problem — security domain experts, product analytics, engineers who understand specific signal pipelines — without maintaining a fixed team that would lack the right