Full Job Description
Own end-to-end projects that bring intelligence into production. You will be responsible for systems that help agents reason, plan and evaluate themselves, with full autonomy to plan projects, define success, run experiments and decide when a system is ready to ship. This role suits engineers who want to operate like researchers and builders at once.
What You Will Do
Design and iterate multi-agent architectures, encoding autonomy boundaries, tool use and fallback behavior that make agents safe and reliable.
Build scalable offline and online evaluation pipelines that run hundreds of experiments, with golden tasks, labeling strategies and metrics.
Instrument the stack to detect regressions, track error taxonomies and drive closed-loop improvement.
Engineer context and retrieval: prompt stacks, indexing pipelines and parsing of messy documents into structured representations.
Scope projects clearly, write concise specs and communicate progress within your pod.
Requirements
What We Are Looking For
Four to twelve years as a machine learning engineer, with end-to-end LLM-based agent applications including benchmarking, orchestration and evaluations.
Experience at a fast-paced startup (Series A to D), a quant or hedge fund, or a top-tier technology company.
Experience running structured ML experiments and building evaluation infrastructure.
A CS, Physics, Math or related technical degree from a top school. The role is applied rather than research-focused.
A very clear communicator who can break complex ideas down.
Currently located in the US or Canada, ideally on the East Coast.
Bonus Points
Deep Python expertise and LLM or transformer-based systems.
Interest in the impact of AI on professional services and the economy.
Benefits
Compensation and Benefits
$175,000-$350,000 base salary plus highly competitive equity, depending on experience. Visa transfers and sponsorship are available.
Location and Work Model
Full-time and on-site five days a week in Flatiron, New York City.
