Founding AI Engineer

pulliAhead Health is building Europe's AI-powered preventive health platform. /liliYou build the AI systems at the core of our product, end-to-end. You help set the technical direction, define the conventions, and shape the paradigms that follow. /liliYou co-own the user experience on AI-driven features alongside design and product. This is not a backend-only role; /liliYou hold a clinical-grade bar on AI quality, reliability, and evaluation discipline. In healthcare every output influences a real decision; /li /ul pYOUR ROLE /p h3Build from the ground up /h3 h3End-to-end AI ownership /h3 pDesign, build, and operate AI systems across the product, from problem framing to production scale. You shape the area that owns them. /p pGround-truth datasets, regression suites, online and offline metrics. No system goes live without a defensible answer to "how do you know it works." /p h3Co-own the AI experience /h3 pPartner with design and product on conversational flows, refusal behaviors, trust signals, and how uncertainty is communicated. Not a backend-only role. /p h3AI that serves the customer's health /h3 pLLMs, agents, retrieval, classical ML, computer vision, medical imaging. The form matters less than the depth — AI here is whatever moves outcomes. /p h3Safety and compliance by design /h3 pDoctors-in-the-loop review, output validation, and guardrails aligned with CE/MDR, EU AI Act, GDPR, and Swiss FADP — design constraints, not checkboxes. /p h3Set the conventions the team adopts /h3 pDefine prompt design, evaluation, and observability practices the rest of the team will follow. Educate on what matters and what doesn't. /p h3Your role /h3 pPersonalized, preventive healthcare has historically been a luxury. AI changes that. Ahead Health raised its Seed round with great investors and is now scaling. /p pAs our Founding AI Engineer, you build the AI systems that power our product, and you shape the area that owns them. The surface is broad: LLM‑driven experiences (personalized health plans, conversational health interfaces, intelligent triage, proactive nudges) on one side; classical ML, computer vision, and statistical modeling on the other (medical imaging analysis, body composition, longitudinal risk and trajectory modeling). AI here is whatever serves the customer's health, not whatever is in vogue. You own these systems end-to-end; from data and context engineering, to model and prompt design, to evaluation, deployment, observability, and continuous improvement. /p pDeploying AI for consumer health carries dimensions you won't find in a typical AI role. You engineer for medical accuracy, graceful handling of uncertainty, and error prevention at every layer. You navigate regulatory frameworks, not as abstract compliance boxes but as design constraints that shape how systems work. You build trust through transparency: making it clear what the AI knows, what it doesn't, and when to involve a doctor. And you balance deep personalization with rigorous data privacy, ensuring that sensitive health data is always handled with the care it demands. /p pThis is not a research-only role. You ship AI to production, with the evaluation and observability discipline that production demands. On AI‑driven features you stay close to the experience itself, not just the system underneath. How AI shows up to customers matters as much as how it works under the hood. You'll work closely with our CEO, CTO, CPO, doctors, designers, and engineers. /p h3Responsibilities /h3 ul libBuild production AI systems. /b Design, build, and operate AI systems end-to-end across our product, from problem framing to production scale. /li libCo‑own UX on AI‑driven features. /b AI is only as useful as how it lands with the customer. Partner with design and product on conversational flows, refusal behaviors, trust signals, and how uncertainty is communicated. /li libOwn AI quality end-to-end. /b Build rigorous evaluation frameworks, automated monitoring, and guardrails so that AI outputs meet clinical‑grade standards. Bring evaluation discipline to every system you ship: ground‑truth datasets, regression suites, online and offline metrics, dashboards, and alerts that catch quality degradation before customers do. No system goes live without a defensible answer to "how do you know it works." /li libArchitect for production. /b Make deliberate choices on prompt and context engineering, agent orchestration, retrieval, model selection, latency and cost. Design fallback paths and graceful degradation so the system stays useful when models fail. /li libEnsure medical accuracy and safety. /b Implement doctors‑in‑the‑loop review systems, output validation, and safety guardrails aligned with regulatory requirements (EU AI Act, GDPR, Swiss FADP). /li libDrive data quality for AI. /b Shape how health data flows into AI systems. Define what data you need, how it's structured, and how feedback loops keep AI outputs improving over time. /li libRaise the AI bar. /b Set the conventions for prompt design, evaluation, and observability that the rest of the team adopts. Evaluate emerging models, frameworks, and techniques; bring relevant innovations to production quickly and pragmatically. Educate the rest of the team on what matters and what doesn't. /li /ul h3Need to have /h3 ul liFive to 8+ years of software engineering experience, with 3+ years building and shipping production AI systems. The form (LLM, classical ML, computer vision, hybrid) matters less than the depth. /li liStrong AI fundamentals across the modern toolkit. LLMs and foundation models (prompt and context engineering, retrieval, agent orchestration), and at least solid grounding in classical ML (training, validation, deployment). Exposure to computer vision, medical imaging models, or fine‑tuning is a strong plus. /li liProduction AI quality discipline. You have shipped AI systems backed by real evaluation infrastructure, and you do not ship without evals. /li liSystems thinking. You make deliberate trade‑offs between accuracy, latency, cost, and complexity. You design for failure modes, not just happy paths. /li liCare about how AI feels to users. You bring UX opinions to AI features and don't hand the UX off to someone else once the model is wired up. /li liFounding temperament. You've built something from zero before. You're comfortable defining practices that don't exist yet, making decisions with limited precedent, and being the most experienced person in the room on building AI features. /li liStrong software fundamentals. Python, API development, testing, CI/CD, cloud (GCP preferred). Enough engineering depth to own a system from prototype to production without depending on a separate platform team. /li liPassionate about responsible AI in health. Deep appreciation for data privacy, model bias, explainability, and the ethical considerations of applying AI to health. /li /ul h3Great to have /h3 ul liYou've built AI‑first products from scratch, as a founder, early‑stage employee, or on a new product line. /li liFine‑tuning, model optimization, distillation, or multi‑agent systems experience. /li liMedical imaging or classical medical ML experience (e.g. nnU‑Net, segmentation, computer vision on DICOM, body composition). /li liHealthcare domain knowledge: clinical workflows, medical terminology, FHIR standards, digital health, or clinical decision support. /li liExperience building AI features in regulated or high‑stakes domains, e.g. healthcare, fintech, insurance, or similar, where output accuracy and safety directly affect end users. /li liResearch background: M.Sc. or Ph.D. in Computer Science, AI/ML, or related field. /li liSwiss or European healthcare compliance experience. /li /ul h3What we offer /h3 ul liFull‑time. Zürich preferred (relocation assistance negotiable); also open to Europe‑based candidates within 1 time‑zone of CET. CHF 145K–155K + significant equity. /li liShape the future of preventive health in Europe through cutting‑edge AI. /li liDirect access to medical experts and real customer feedback to inform your work. /li liSmall, experienced, technical, and mission‑led team from Google, Coolblue, Kaia Health, and Oxford. /li /ul /p #J-18808-Ljbffr