Staff Machine Learning Engineer

Altana•New York, NY

1d•$205,000 - $240,000•Hybrid

About The Position

Altana is seeking a Senior Staff Machine Learning Engineer to lead the AI behind Altana Nexus, its agentic platform for global trade. The role focuses on public and proprietary trade data—including shipments, customs filings, and trade documents—and on building models and AI agents that support real-time decisions about risk, compliance, and enforcement for governments and large companies. The engineer will set technical direction for agentic systems and lead reliable, cost-efficient systems spanning data ingestion, document understanding, entity resolution, inference, and model routing. They will adapt and post-train models using analyst feedback, production outcomes, and agent traces; develop evaluation, tracing, and feedback systems; mentor senior engineers; and collaborate with engineering, data science, product, and customers to turn ambiguous problems into shipped capabilities. The role is based in Brooklyn, New York, with hybrid work flexibility. Candidates must live within commuting distance of Altana’s Williamsburg office. The stated annual base salary range is $205,000–$240,000 USD. Full-time offers generally include competitive equity grants, and compensation may also include a discretionary bonus or variable compensation plan, depending on the role. Altana offers benefits including flexible time off, paid parental leave, medical, dental and vision coverage, supplemental insurance, a 401(k) program, commuter benefits, wellness resources, pet insurance, an Employee Assistance Program, and a Dependent Care FSA.

Requirements

Bachelor’s or master’s degree in Computer Science, Machine Learning, or a related field, or equivalent experience.
At least 8 years of experience building and shipping production ML systems, with a track record of product-oriented ML solutions, including recent LLM-powered products.
Deep expertise in PyTorch and the modern Python ML stack, including Hugging Face, scikit-learn, NumPy, and Pandas or Polars.
Hands-on experience building production LLM applications using retrieval-augmented generation, tool or function calling, structured outputs, and agentic workflows.
Experience shipping production agentic systems and understanding their failure modes, including tool use, planning, state, recovery, and human-in-the-loop processes.
Experience post-training or adapting LLMs for production use cases using methods such as SFT, DPO, RLHF/RLAIF, or reward modeling.
Experience building rigorous evaluation suites, golden datasets, and LLM-as-judge pipelines.
Sound judgment in selecting frontier-model APIs, fine-tuned open-weight models, or classical ML models while balancing cost, latency, and accuracy.
Willingness to work directly with data and domain details to apply advanced techniques to real-world problems and achieve measurable outcomes.
Commitment to engineering excellence, supportability, and knowledge-sharing.
Strong written and verbal communication skills.
Daily use of AI coding tools such as Cursor or Claude Code.
Must be within commuting distance of the Williamsburg, Brooklyn office.

Nice To Haves

Experience building multi-agent systems and agent tooling, such as LangGraph.
Experience with hybrid retrieval, including vector search, re-rankers, and embedding fine-tuning.
Experience with vision-language models for document understanding, such as invoices, bills of lading, and customs forms.
Experience in multilingual NLP.
Experience with LLMOps tools such as MLflow and Databricks Mosaic AI, and with guardrails.
Experience with AI safety, guardrails, or auditability in regulated or high-stakes fields such as customs, legal, healthcare, or finance.

Responsibilities

Drive the technical direction, architecture, and roadmap for agentic systems that process incoming trade data and generate real-time risk, classification, and compliance decisions.
Lead the design of reliable, cost-efficient systems for trade-data ingestion, document understanding, entity resolution, inference, and routing across models.
Fine-tune and post-train models using analyst feedback, production outcomes, and agent traces, applying techniques such as SFT, DPO, and reinforcement learning with domain graders when they outperform prompting measurably.
Build evaluation, tracing, and feedback systems that inform release decisions and catch production regressions.
Mentor senior engineers, raise the technical bar across Engineering, and help shape hiring and leveling for machine-learning talent.
Collaborate with software engineers, data scientists, product managers, and customers to develop business ideas and turn ambiguous problems into shipped capabilities.

Benefits

Flexible Time Off policy.
Paid parental leave: 14 weeks for non-birthing, adoptive, and foster parents, and up to 26 weeks for birthing parents, paid at 100% of base salary.
Medical, dental, and vision benefits with employer contributions; the high-deductible health plan is fully employer-paid for employees, with an employer contribution to a Health Savings Account.
Flexible Spending Account option.
Life, short-term disability, long-term disability, and AD&D insurance coverage at no cost to employees.
Fidelity 401(k) retirement savings program.
Pre-tax commuter benefits for public transit or parking.
Free premium Calm subscription.
Pet insurance and a pet veterinary service and telehealth discount plan.
Free access to a confidential Employee Assistance Program.
Dependent Care Flexible Spending Account.
Competitive equity grants included in the majority of full-time offers.
Depending on the role, a discretionary bonus or variable compensation plan.