The Opportunity
We are looking for a Full-Stack AI Engineer, Clinical Systems to help build the intelligence layer at the center of the Alnu Health platform.
You will own systems that transform clinically reviewed knowledge and longitudinal patient context into reliable, personalized support. This includes coach evaluation, patient memory, knowledge ingestion and retrieval, model orchestration, validation, observability, and the backend services connecting those capabilities to our patient and clinician experiences.
This is a deeply technical and highly product-oriented role. You might spend one day designing an evaluation framework in Python, the next improving how patient memory is structured and retrieved, and the next shipping that capability into our mobile app or clinician portal.
You will work directly with engineering leadership, product leaders, and clinical experts, with meaningful autonomy over systems central to Alnu’s product and long-term technical direction.
How We Think About AI
Most AI job descriptions sound the same. This role is different because our philosophy is different.
We do not believe the future of healthcare will be built by maximizing tokens, choosing the largest available model, or wrapping a chat interface around an API.
We also do not hire engineers because they are experts in one particular model, provider, or framework.
AI is evolving too quickly to build a company around Claude, GPT, Gemini, Qwen, or whichever model is released next. We continuously evaluate models, architectures, retrieval strategies, orchestration patterns, and supporting infrastructure. When something better becomes available, we should be able to assess it, adopt it, combine it with other systems, or replace what came before.
Models are tools. Judgment is the advantage.
We believe trustworthy clinical AI requires an engineered system:
- Clinically reviewed, evidence-grounded, and source-traceable knowledge
- Longitudinal memory that understands a patient over time
- Retrieval that prioritizes clinical relevance, not semantic proximity alone
- Evaluation systems that measure quality, safety, consistency, and personalization
- Deterministic constraints and validation around model behavior
- Clear observability into which knowledge, memories, tools, and rules influenced an output
- Deliberate optimization of model choice, latency, reliability, and cost
- Product experiences that strengthen the relationship between patients and clinicians
The objective is not to generate more. It is to deliver the most useful, grounded, and appropriate support for a specific patient at the moment they need it.
AI should extend clinicians, not replace them.
Every engineering decision should ultimately make care more effective for patients or make it easier for clinicians to deliver that care.
What You’ll Do
Clinical AI Systems
- Build a repeatable evaluation system for Alnu Health’s user-facing AI companion using quality, safety, personalization, consistency, and clinical grounding
- Improve how longitudinal patient context is captured, structured, retrieved, prioritized, and used
- Strengthen the pipeline through which clinical guidelines, expert protocols, and reviewed source material become versioned and testable system knowledge
- Develop and improve retrieval, orchestration, verification, and constraint systems
- Design feedback loops that turn product usage and expert review into measurable system improvements
- Optimize model usage for quality, latency, reliability, and cost
Full-Stack Product Engineering
- Design and build production backend services using Python and TypeScript
- Ship AI-enabled capabilities across our patient-facing companion, clinician portal, APIs, and supporting infrastructure
- Contribute across the full application stack, including APIs, web interfaces, and mobile experiences
- Investigate failures across data, retrieval, prompts, models, business logic, and product experience
- Write clear, maintainable code and ship through small, well-tested releases
Clinical Quality and Collaboration
- Work with clinicians to understand the intent behind guidance and encode that intent in reliable systems
- Treat patient privacy, security, auditability, and clinical safety as core engineering requirements
- Communicate technical tradeoffs clearly across engineering, product, clinical, and company leadership
Who You Are
- Roughly 3 to 5 years of professional software engineering experience, or equivalent evidence of technical depth and end-to-end ownership
- Highly proficient in Python, TypeScript, and JavaScript
- Experienced building and operating production backend systems
- Able to ship meaningful product features across more than one layer of the stack
- Experienced with production LLM-enabled systems beyond basic API integration
- Understand the strengths and limitations of RAG, embeddings, structured retrieval, tool use, and model orchestration
- Have built evaluations, test harnesses, feedback systems, or observability for nondeterministic software
- Care about maintainability, reliability, latency, and cost alongside model quality
- Can join an unfamiliar codebase, map the system, ask focused questions, and begin contributing quickly
- Comfortable working from an incomplete specification and owning the design of a feature from initial ticket to production deployment
- Want substantial ownership in an early-stage startup where engineers participate in product, architecture, testing, and deployment
- Motivated by improving outcomes for patients with chronic disease
We care more about evidence of judgment, ownership, and shipped systems than pedigree or a perfectly matched job title.
Particularly Valuable Experience
- Healthcare, digital health, clinical data, or other high-stakes software
- Clinical knowledge bases, ontologies, retrieval systems, or longitudinal data
- LLM evaluation, fine-tuning, or model behavior analysis
- Agentic systems, tool orchestration, verification layers, or constrained generation
- React, React Native, or consumer mobile products
- HIPAA, SOC 2, privacy engineering, or regulated environments
- Wearable, laboratory, EHR, or third-party health data integrations
You do not need to arrive as an expert in medicine. You do need the ability and work ethic to learn a high-stakes domain carefully and collaborate closely with the people who practice it.