Software Engineer, AI Agent Governance

Every breach, every failed audit, every new hire stuck in onboarding purgatory traces back to the same question: who has access to what — and should they?

Identity is where modern companies break. Too much access leaks data. Too little grinds the business to a halt. Lumos is the industry's first Autonomous Identity Platform — software that doesn't just surface access risk, but acts on it. We started with a self-service AppStore and grew into the critical infrastructure connecting apps, identities, and data for GitHub, MongoDB, Mars, Pinterest, Anduril, Intercom, Roku, and Netskope. At the center is Albus, our AI agent for identity — mining roles, approving access, and catching the drift humans miss.

You'll ship work that thousands of IT and security teams touch every day. You'll own problems without playbooks, because the category is being redefined in real time — by us.

Why Lumos?

• 🚀 Jump on a rocketship. Since launching out of stealth in April 2022, we've grown from 20 to ~130 Alchemists and raised over $65M from a16z, Scale, Neo, Greg Brockman (President at OpenAI), and Phil Venables (CISO at Google).
• ⭐ Customers love us — and so does the industry. Named a Strong Performer in the 2026 Gartner Voice of the Customer for IGA — 4.7/5, with 96% of reviewers saying they'd recommend us. Our customers cut time-to-access by 98%, slash IT tickets by 72%, and eliminate excess privilege at scale.
• 🏆 A culture worth joining. Three years running on Forbes' Top 100 Best Startup Employers in America (2024, 2025, 2026) — measured on culture, retention, growth, and job creation. That's our Alchemists, not our marketing.

AI agents are on every company laptop. They are on more cloud workloads each week. The numbers are going up and to the right, fast. These agents act with whatever access the person running them has. Security teams can't see which tools their agents call, let alone approve or block them. Lumos is building the governance layer for these agents, and we're rolling it out to customers right now.

As a Software Engineer on AI Agent Governance at Lumos, you have massive opportunity to help build a brand new product from the ground up in one of the fastest moving markets in software. You'll join a small team who ships fast.

✨ Your Responsibilities

• Ship backend features end to end, from a rough customer ask to running in production, often in days rather than weeks.
• You don’t just ship to a spec, rather you ensure value is delivered to customers. You are a valued member of the oncall rotation.
• Build the testing systems and harnesses that let agents ship changes safely.
• Work directly with customers when issues arise.
• You utilize agents heavily in your day to day work, but you have a deep understanding of the system architecture and can whiteboard it for a coworker at a moment’s notice.

🙌 What We’re Looking For

• 1+ years of professional software engineering experience, or an equivalent track record of shipping real products.
• Strong backend fundamentals. You've built APIs, databases, and services that run in production. Our stack is Python, Postgres, OpenSearch, and gRPC on AWS.
• You're AI native. You understand how agents work under the hood: tool calls, MCP servers, hooks, and context.
• You're comfortable with ambiguity. You've worked somewhere small and fast, you start before the spec is finished, and you change direction when customers tell you something new.
• You take ownership. This is a high-intensity role on a product that's still finding product market fit. We work hard and move fast, and we want people who are excited by that.
• Nice to have: experience with endpoint or client software (MDM, software that runs on laptops), a security or identity background.

⭐️ What We Value

We also care about whether you would be a good fit for Lumos based on the values and characteristics that define how we achieve outcomes, not just your resume.