we are building an agentic development lifecycle, an infrastructure of autonomous agents that work alongside our engineers to accelerate and improve how we build software. Our goal is to ship faster, with higher quality, and to continuously tighten the feedback loop between what the agents produce and what engineering actually needs. Over time, this system should compound: every improvement makes the next one easier to reach.
we are an enterprise secure browser used by some of the largest organizations in the world. It's a complex, multidisciplinary product spanning browser core, frontend, extensions, and backend services, and it runs at scale for customers who need it to always work. The bar for what we ship is high. That means whatever agentic infrastructure we build has to meet the same standard. We're not here to vibe code our way to production.
We're looking for an AI Engineer with a product builder's mindset. You have real experience with AI and agentic workflows, and you know how to take a complex project from idea to adoption, technically and organizationally. That means working across teams, aligning with security, infrastructure, and other engineering groups, and understanding that building the system is only half the job. Getting people to trust it is the other half.
We aren't looking for a conventional senior developer; we need someone whose mindset is adapted to technical challenges that didn't even exist 18 months ago.
Requirements: Your Impact
Design and implement automated evaluation loops, static analysis, and rigorous quality gates to ensure the ADLC process doesn't just write code, but consistently produces great, production-ready code.
Help the team tackle complex, hard problems to elevate our autonomous development product from "good" to "excellent".
Lead complex initiatives in Context Engineering and Prompt Engineering.
Manage and orchestrate the complex ecosystem of autonomous agents utilized for internal development.
Serve as a leading individual in a very strong team professionally and personally - Were looking for someone who not only delivers his own work but improves that of those around them.
Find space for growth to push the entire team or group forward - New projects, changing processes or improving existing tools.
View prompt engineering as a core engineering discipline-where rewriting agent behavior is a versioned, reviewed, and tested code change.
Act with a debugging temperament; conduct deep-dive analyses of raw agent transcripts to diagnose non-deterministic failures and ascertain root causes instead of merely working around them.
Your Experience
At least 8+ years of experience in software development, architecture, or owning operational systems in production.
Computer Science B.Sc. or equivalent education or equivalent military experience required.
A product builder's mindset: you can extract requirements, talk to stakeholders, and tell the difference between what's important and what's noise.
Experience in building production grade agents. Deep understanding of the agent loop, its states and transitions. You know how to build it correctly, not just use it.
Positive can-do mindset, able to work independently and within a team.
Hands-on experience with LLM APIs, including a practical, highly-skeptical understanding of token costs, caching, context windows, and model failure points.
You know how to build the right context for a task, including memory systems, session storage, and vector databases.
You understand where LLMs fail and how to design around those failure points.
You've used traces or observability tooling to diagnose and improve agent behavior.
A systems-level background that touches reliability, observability, or platform engineering, with a strong preference for writing narrow, deterministic code over building hypothetical abstractions.
Experience in the cybersecurity space - an advantage.
This position is open to all candidates.