This role owns the fullstack experiences where clinicians and operators meet Hopper’s AI. That includes the conversational and agentic interfaces (chat, voice, streaming, and tool-calling) that let people interrogate clinical data and drive workflows in natural language, along with the AI-powered features woven directly into HopperOS. The developer turns model output into interfaces that are fast, legible, and trustworthy at the point of care, building across the full stack against Hopper’s clinical data contracts.
ESSENTIAL FUNCTIONS
- Conversational & Agentic AI Interfaces
- Own the agent experience across chat and voice interaction, streaming responses, conversation history, and model-generated titles and summaries.
- Make model reasoning legible by surfacing tool calls, citations, intermediate steps, and uncertainty.
- Build low-latency, real-time experiences over WebSockets and streaming APIs that stay responsive under clinical conditions.
- Render dynamic model output into on-the-fly charts, tables, and structured clinical summaries.
- Partner with the CIE and orchestration team to expose tiered model routing, skills, and reasoning in ways users can trust and act on.
- AI-Powered Clinical Product
- Embed AI into HopperOS workflows, including documentation and ambient-scribe surfaces, inline suggestions, and natural-language query over clinical and operational data.
- Build against Hopper’s data contracts, including the FHIR API surface and Athena-integrated clinical data.
- Ship the rich, interactive product surfaces where interactivity clearly adds value.
- Craft, Evaluation & Iteration
- Own fullstack quality for AI features, including streaming UX, performance, accessibility, and responsiveness across desktop and point-of-care tablet and mobile.
- Close the loop with evals by measuring the quality of generated UI and agent responses.
- Work forward-deployed with real clinician and operations users, incorporating feedback into design and build-out.
EDUCATION & CERTIFICATIONS
- Bachelor’s degree in Computer Science, Software Engineering, or a related technical field. Equivalent professional experience will be considered.
- Certifications in cloud, AI/ML, or modern web development preferred.
EXPERIENCE
- 5+ years building and shipping production web frontends or fullstack products, ideally with 0-to-1 ownership.
- Hands-on AI product experience building chat or agent UIs, tool-calling, streaming responses, and app-layer context and prompt management.
- Strong frontend craft across component-driven UI, state management, and real-time interfaces (WebSockets, streaming); experience with React is a plus, though craft and adaptability matter more than any specific framework.
- Experience turning structured and unstructured model output into UI that is dynamic, legible, and trustworthy.
- Healthcare or clinical software experience, including familiarity with FHIR/HL7 or EHR integration (e.g., Athena), preferred.
KNOWLEDGE
- Systems-level UX judgment, with the ability to reason about latency, failure modes, and safety in AI experiences.
- Familiarity with agent architectures and tool-calling systems end to end.
- Knowledge of LLM evaluation tooling (e.g., Braintrust) and voice interfaces.
- Understanding of ML and LLM model development; curiosity about Elixir/Phoenix or computer vision a plus.
SKILLS & ABILITIES
- Strong product instincts, comfortable prototyping live with users and shipping feedback within days.
- A feel for the model boundary between raw model output and trustworthy, legible UI.
- Strong collaboration across clinical, product, and engineering stakeholders.
- Adaptability to work in a fast-paced, greenfield environment and go deep in whatever stack the problem needs.
ENVIRONMENTAL WORKING CONDITIONS
- Normal office environment
- May require some travel
PHYSICAL/MENTAL DEMANDS
- Requires sitting and standing associated with a normal office environment.
- Manual dexterity using a calculator and computer keyboard.
ORGANIZATIONAL REQUIREMENTS
- HOPCo Mission, Vision and Values must be acknowledged and adhered to