Member of Technical Staff - Applied ML

Member of Technical Staff - Applied ML

Company: Basis
Location: New York, NY (Flatiron office, in person 5 days per week)
Compensation: $175,000 - $350,000 + highly competitive equity
Employment Type: Full-time
Visa Sponsorship: Visa transfers; can sponsor all types

About Basis

Basis started from the belief that AI agents would become integral to knowledge work, and that accounting (structured, high-stakes and essential to every business) would be among the first domains transformed. Three years in, Basis can complete a partnership tax workbook end to end, and accountants use it daily to create complex journal entries and debug reconciliations, with capabilities improving every month.

Basis has raised a $140M Series B.

The Role

As an ML Engineer at Basis, you will own end-to-end projects that bring intelligence into production: the systems that help its agents reason, plan and evaluate themselves. You will have full autonomy to plan projects, define success, run experiments and decide when a system is ready to ship. This is an applied role for engineers who want to operate as researchers and builders at once.

What You Will Do

• Design and iterate multi-agent architectures that automate real accounting workflows, with clear autonomy boundaries, tool usage and fallback behavior.
• Manage context and memory across agent steps; route, evaluate and optimize models under latency, cost and accuracy constraints.
• Build scalable offline and online evaluation pipelines that run hundreds of experiments automatically, with golden tasks, labeling strategies and metrics.
• Instrument the stack to catch regressions, track error taxonomies and drive closed-loop improvement.
• Architect prompt stacks, retrieval and indexing pipelines, and document parsing into structured representations agents can reason about.
• Scope projects with concise specs, build and test end to end, and communicate progress clearly within your pod.

What You Bring

• 4-12 years as a machine learning engineer
• Experience at a fast-paced startup (Series A-D), a tier-1 tech company or a hedge fund
• End-to-end LLM agent applications: benchmarking, model orchestration and evals for agent behavior and reliability
• Structured ML experimentation: framing hypotheses, building evaluation infrastructure, iterating on measurable results
• CS, physics, math or other technical degree from a top school
• Very clear communication
• Located in the US or Canada and able to work in the New York office 5 days a week

Nice to Have

• ML products and underlying models at a fast-paced company
• Deep Python and LLM/transformer expertise
• Interest in AI's impact on accounting and finance

Interview Process

Initial screen with leadership, meet-the-team screen, technical coding interview, onsite plus references.

Tech Stack

Python, Postgres, LLMs, agent frameworks, evaluation pipelines

REVENUE: 21% of first-year salary. Est. fee per hire $37K-$74K; 10 seat(s) = up to $551K if all filled.

TARGET COMPANIES (client (DeepMind/Ramp/Databricks named) + suggested): Ramp, Databricks, Google DeepMind, Harvey, Hebbia, Citadel, Jane Street.

BEST-FIT CANDIDATE: 4-12 yrs; end-to-end LLM agent systems + evals; 0-to-1 ownership at startup/tier-1 tech/hedge fund; quantified production impact; visa: transfers; can sponsor; location: NYC 5 days. Applied not research: generally avoid Masters/PhD-heavy academic profiles and big-company-only backgrounds.