About Altitude
Altitude is building the clinical AI execution platform for risk-bearing care organizations. We help care teams Perform — getting every primary care clinician to specialist-level decisions in the room, and executing the operational follow-through that gets chronic patients to goal faster. Where organizations need more capacity, we Extend it — delivering care directly, under their brand and rules, to the patients they can't reach.
Underneath both is a network of agents, not a single assistant, working the same longitudinal patient context graph from both sides of the visit. On the clinician side, condition-specific subagents continuously read the chart, extract clinical signal, and generate structured recommendations through our in-EHR conversational agent. On the patient side, agents reach out directly, gather updates, and triage concerns. The two sides hand off to each other in real time: a patient agent escalates the moment something needs judgment, a clinician agent (or a human) picks up with full context, and control passes back once the visit's follow-through is handled.
We believe AI can meaningfully improve how clinical work gets done, but only when it is developed with rigorous evaluation, deep observability, and a clear understanding of the clinicians and patients it serves.
We are supported by top-tier VCs in healthcare and AI, including Lerer Hippeau, AlleyCorp, Tau Ventures, Chrysalis, MVP Ventures and Sequel. We are a small, product-driven team working at the intersection of software, AI, and healthcare.
The role
We’re looking for an AI Product Engineer to help build the technical foundation for our clinical AI products.
Your primary focus will be our agentic clinical platform: an environment that gives AI systems the context, tools, feedback, and safeguards required to complete complex clinical workflows. You will improve how these systems use tools, manage context, recover from failures, and collaborate with people.
You will own meaningful product and platform capabilities from early prototype through production, including the evaluations, observability, testing, and feedback loops.
This is a hands-on product engineering role for someone who moves comfortably between product thinking, full-stack development, and applied AI. You should enjoy working from ambiguous problems, testing ideas quickly, and turning the strongest prototypes into reliable software.
You’ll work in person with our team in New York City.
What you’ll do
- Build and evolve our agentic clinical platform, including context management, tool use, orchestration, permissions, memory, and human-in-the-loop workflows.
- Prototype new clinical AI experiences and rapidly turn promising ideas into production-quality software.
- Design evaluations that measure task success, clinical quality, reliability, safety, latency, and cost.
- Create repeatable test environments, representative clinical scenarios, regression suites, and launch criteria for AI-powered features.
- Build observability across model behavior, tool calls, traces, errors, user interventions, and end-to-end workflow outcomes.
- Develop feedback loops that turn production behavior and expert review into better prompts, tools, evaluations, and product decisions.
- Explore new models, agent frameworks, inference techniques, and developer tools; run disciplined experiments and identify what is genuinely useful for our product.
- Work across the stack, from model integrations and backend services to internal tools and user-facing product surfaces.
- Partner closely with clinicians, product leaders, designers, and engineers to translate real clinical workflows into thoughtful software.
- Make pragmatic tradeoffs among speed, reliability, clinical risk, privacy, security, maintainability, and user experience.
- Help establish the engineering practices and technical standards for building trustworthy clinical AI systems.
What we’re looking for
- 5+ years of professional software engineering experience, or equivalent evidence that you can independently build and ship high-quality products.
- Strong product judgment and a track record of turning ambiguous ideas into useful software.
- Fluency in Python, TypeScript, or similar languages, with the ability to work across backend systems, APIs, data models, and product interfaces.
- Hands-on experience building with LLMs or agentic systems, including tool calling, structured outputs, retrieval, context management, or multi-step workflows.
- Experience designing tests, evaluations, telemetry, or monitoring for systems whose behavior is probabilistic rather than fully deterministic.
- The ability to move quickly during exploration and then harden the right ideas for production.
- Strong debugging instincts across prompts, model behavior, application code, data, infrastructure, and user workflows.
- Clear written and verbal communication, particularly when explaining technical behavior and tradeoffs to clinical or non-technical collaborators.
- Intellectual curiosity and a habit of staying current with rapidly evolving models, research, tools, and engineering practices.
- A willingness to work in person in New York City.
You might be especially well suited if
- You have been an early engineer, previous founder, or product-minded full-stack engineer in a fast-moving environment.
- You have built agent harnesses, coding-agent infrastructure, developer tools, workflow engines, or systems for supervising autonomous software.
- You have experience with LLM evaluations, trace analysis, prompt and tool regression testing, model comparison, or production AI observability.
- You have worked in healthcare or another high-stakes domain where accuracy, privacy, auditability, and human oversight matter.
- You enjoy talking directly with domain experts and converting messy real-world workflows into clear product and technical abstractions.
- You are skeptical of impressive demos until they survive realistic evaluations and production use.
- You do not need a research degree or experience training foundation models. We care more about whether you can understand model behavior, build excellent software around it, and create the systems required to improve it.
What success looks like
In your first six months, you will have:
- Shipped meaningful improvements to the clinical harness that expand what our AI systems can accomplish safely and reliably.
- Established evaluation suites and regression tests that the team trusts when making launch decisions.
- Made model and agent behavior substantially easier to observe, debug, and improve.
- Taken multiple ideas from rough prototype to real use by clinicians or internal teams.
- Helped the company distinguish genuine product progress from improvements that only look good in a demo.
Interested?
Email us at [email protected] with your resume and a quick blurb on why this role might be a good fit for you