The role
We’re hiring a Forward Deployed Engineering Manager to lead the design, development, and delivery of reinforcement learning environments for agentic AI systems.
You’ll manage a team responsible for building sandboxed, reproducible environments—terminal-based workflows, browser automation, and computer-use simulations—that power both model training and human-in-the-loop evaluation. This is a hands-on leadership role where you’ll set technical direction, guide execution, and stay close to architecture and critical systems.
What You’ll Do
- Lead, hire, and develop a high-performing team of Forward Deployed Engineers, setting a high bar for ownership, velocity, and technical quality
- Own the RL environment roadmap, aligning team execution with customer needs and evolving model capabilities
- Oversee development of sandboxed environments (terminal, browser, tool-augmented workspaces) that support deterministic execution and multi-step agent interaction
- Ensure reliability, observability, and data integrity through strong instrumentation (logging, trajectory capture, state snapshotting)
- Drive infrastructure excellence across containerization, sandboxing, CI/CD, automated testing, and monitoring
- Partner cross-functionally with data operations, product, and leading AI labs to define task design, evaluation protocols, and environment requirements
- Enable rapid prototyping and iteration, helping the team move from ambiguous requirements to production-ready systems quickly
- Stay close to the technical details—reviewing architecture, unblocking complex issues, and guiding design decisions
What We’re Looking For
Required
- 5+ years of software engineering experience (Python)
- 2+ years of experience managing or leading engineers in fast-paced environments
- Strong experience with containerization and sandboxing (Docker, Firecracker, or similar)
- Solid understanding of reinforcement learning fundamentals (MDPs, reward design, episode structure, observation/action spaces)
- Background in infrastructure, developer tooling, or distributed systems
- Strong debugging skills and systems thinking across layered, containerized environments
- Ability to operate in ambiguity and translate loosely defined problems into clear execution plans
- Excellent communication and stakeholder management skills
Preferred
- Experience building or working with RL environments (Gym, PettingZoo) or agent benchmarks (SWE-bench, WebArena, OSWorld, TerminalBench)
- Familiarity with cloud infrastructure (GCP or AWS)
- Prior experience in AI/ML platforms, data companies, or research environments
- Contributions to open-source projects in RL, agents, or developer tooling
Why This Role Matters
RL environment quality is a critical bottleneck in advancing agentic AI. Poorly designed or unreliable environments introduce noise into training loops and directly impact model performance.
In this role, you’ll lead the team building the environments that define how models learn—working across a range of cutting-edge projects with leading AI labs. Alignerr offers the speed and ownership of a startup with the scale and resources of Labelbox, giving you the opportunity to have outsized impact on the future of AI.
About Alignerr
Alignerr is Labelbox’s human data organization, powering next-generation AI through high-quality training data, reinforcement learning environments, and evaluation systems. We partner directly with leading AI labs to build the data and infrastructure that push model capabilities forward.
