Senior Applied AI Engineer (Forward Deployed) [NYC or SF]

Description

The company builds a platform that health systems use to create their own clinical AI agents.

Beneath it sits the Health World Model, which merges a health system's data into a continuously updated, event-level record of each patient's care. Agents built on it cover tasks from intake and scheduling through medication management and clinical decision support.

The agents have handled over 10 million patient encounters and are expected to grow tenfold this year. The platform runs natively on Databricks and integrates with every major EHR. The company has about 35 people, is backed by Madrona, General Catalyst, and Optum Ventures, and works in person in New York and San Francisco.

This role sits in Applied AI, where the platform meets real customer environments. The core engineering team builds the platform, and you take it into health systems, building what each customer needs on top of it as production software.

You own a full customer use case from brief to production, making its architecture decisions, setting its timeline, and answering for the result. Success is measured by the customer's numbers, such as more intake handled, fewer denials, and faster billing close, not by uptime alone.

When a pattern repeats across customers, you spot it and turn it into part of the platform. Applied AI work runs ahead of the platform and shapes its direction.

Career path runs from Senior to Staff or Tech Lead on the Applied AI ladder, based on the scope of work you can be trusted with.

The team judges every decision by whether it improves patients' lives. It moves quickly where it can and slows down where a choice is hard to reverse. Each metric has an owner, a threshold, and a planned response, and metrics nobody acts on get dropped.

Responsibilities

• Own a use case from the deployment brief through production and ongoing operation. Set its internal design, commit to timelines, and ship without anyone needing to review your work behind you.
• Build agent systems for clinical use: layered hierarchies of agents and skills rather than one large model call, with evals that measure output distributions. Nothing reaches a patient until it passes the safety floor and the eval gate.
• Work through integration problems with EHRs, legacy systems, and varied customer data, and find the route from what the customer has to what the platform needs.
• Turn proven patterns into platform capabilities. Move settled work out of agent reasoning and into deterministic data lookups, collapsing steps such as twenty tool calls into one, so the company scales without becoming a services shop.
• Raise problems early, including unvetted specs, unsupported integrations, and scope changes mid-build, before the build absorbs the cost.
• Mentor engineers on your use case and help shape how the team works. Write documentation clear enough for a coding agent or a new hire to continue your work.

Requirements

• Healthcare background: EHR integrations (Epic, Oracle Health, athenahealth), HL7, FHIR, or clinical workflows.
• Experience with Databricks, voice agents, or forward deployed or customer engineering.
• Time at an early-stage startup.
• Experience owning complex production systems that real users depend on.
• Hands-on work with LLMs and agents in production, or the depth to get there quickly.
• Rigor with non-deterministic systems: a few good runs of an agent prove nothing.
• Ability to weigh what a decision costs now against what it makes easier or harder for the team later.
• Comfort with customers and clinicians. You learn the workflow rather than waiting for a perfect spec.
• Craft: your systems fail loudly to you before they fail for a customer, and they keep running while you are away.
• Low ego: the best idea wins regardless of who had it. Direct: you name hard issues, challenge ideas, and commit fully once decided. High agency: you fix what is broken instead of filing tickets. A high personal bar. Skepticism toward rules that lack a reason.

Conditions

Work in person in New York or San Francisco.

Early in the week is for collaborative in-office sprints, Thursday is a protected focus day, and Friday is for consolidating what the team learned.

Benefits

• Broad health, dental, and vision insurance coverage.
• Lunch and dinner catered daily.
• Access to mental health support and wellness coaching.
• A flexible stipend for wellness spending on fitness, therapy, or personal growth.
• A yearly learning budget for courses, books, or conferences.
• A budget to attend conferences for professional development.
• An annual team offsite.
• Opportunities for academic collaboration.
• Unlimited PTO.