Our client, building AI agents that do real accounting work, is looking for a Member of Technical Staff on applied ML in New York.They believe agents will become integral to knowledge work and that accounting, being structured, high stakes and essential to every business, is where that happens first. Three years in, their platform can complete a partnership tax workbook end to end, and accountants use it daily to create complex journal entries, debug reconciliations and prepare technical accounting memos. This role owns end to end projects that bring intelligence into production, with full autonomy to plan, define success and decide when a system ships. In person in Flatiron, five days a week. Salary is $175K to $300K plus highly competitive equity, with visa sponsorship available.Role & Responsibilities:- Design and iterate multi agent architectures that automate real accounting workflows- Encode autonomy boundaries, tool usage and fallback behaviours that keep agents safe and reliable- Manage context and memory for coherence across steps, and route, evaluate and optimise models against latency, cost and accuracy- Build scalable offline and online evaluation pipelines running hundreds of experiments automatically, with golden tasks and clear metrics- Architect prompt stacks, retrieval and indexing pipelines, and parse messy documents into structured representations- Scope your own projects, write concise specs and architecture docs, and instrument systems end to endSkills & Experience:- 3 to 10 years developing complex, data heavy products and machine learning models- Experience building agent systems in production, including tool use, memory and guardrails- Strong evaluation and experimentation discipline, using data rather than intuition to drive decisions- Retrieval and document parsing experience, turning messy inputs into structured representations- Python and Postgres, with the ability to operate like a researcher and a builder at once- Comfortable with full autonomy and in person work five days a week