Senior AI/ML Engineer (Applied, Production)

Company Description

Longevitix is building the operating system for longevity and preventive medicine. Our platform enables physicians to deliver evidence-based, personalized and preventive care at scale by aggregating complex clinical data, translating it into actionable insights, and automating adherence programs that improve long-term outcomes.

We work with concierge, functional, integrative, precision medicine and preventive-focused clinics that demand medical-grade rigor, explainability, and trust. Our mission is to empower physicians with AI that augments clinical judgment rather than replacing it.

We’re hiring a Senior AI/ML Engineer (Applied, Production) to build the AI/ML and LLM “engine” behind this experience: the workflows that transform messy, multi-source health data into structured insight, trustworthy narratives, and scalable clinical enablement—while keeping outputs informational/educational and supportive of clinician judgment.

Role Description

We are seeking a highly skilled Senior AI/ML Engineer to join our team in a hybrid full-time capacity, based in Tel Aviv. In this role, you will design, develop, and implement advanced machine learning models for real-world healthcare applications, with a focus on predictive diagnostics and personalized intervention systems. Your responsibilities will include:

• Own production AI/ML delivery: take features from concept → prototype → production launch → monitoring → iteration.
• Design and implement ML workflows: data pipelines, feature generation, training/evaluation loops, and inference services.
• Build reliable LLM capabilities where relevant: retrieval/grounding, context management, tool-use workflows, evaluation and guardrails.
• Define metrics and run experiments: establish baselines, offline/online evaluation, and clear success criteria.
• Integrate into the platform: ship via APIs and services; ensure traceability, observability, and operational robustness.
• Partner with product + clinical stakeholders: translate needs into measurable AI initiatives while keeping outputs informational/educational.

Qualifications

• Demonstrated track record of deploying ML features to production (multiple shipped projects), including:

- Problem framing, baselines, and model selection

- Evaluation methodology and metric design

- Deployment patterns (batch/streaming/online inference)

- Monitoring, drift/performance tracking, and iteration

• Deep understanding of applied ML (supervised/unsupervised fundamentals, feature engineering, error analysis).
• Strong software engineering skills in Python (testing, architecture, performance, reliability).
• Ability to design systems end-to-end: data → model → service → product experience.
• Comfortable working in a fast-moving startup environment with high ownership and ambiguous problems.

Technical Score (our stack)

• Python (core services, pipelines, ML/LLM integration)
• TypeScript/JavaScript (integration points, shared types, occasional full-stack work)
• Postgres (schema design, performance, migrations)
• AWS (production deployment, security-minded architecture)
• Terraform (infra-as-code)
• APIs (service boundaries, versioning, auth)
• LLMs / GenAI toolchain

- Model APIs (OpenAI/Anthropic/Gemini or equivalents)

- RAG components (embedding models, vector search, rerankers)

- Agent/tool-calling patterns (function calling, workflow orchestration)

- Prompt/version management and safe templating (e.g., Jinja-style)

- Evaluation: golden sets, regression tests, judge models, offline/online eval

- Observability: traces/spans for LLM calls, token/cost/latency tracking, prompt+response

logging (with PHI-safe controls), A/B experiments, alerting

Bonus Points

• Production experience with LLMs: RAG, tool/agent workflows, prompt/version management, evaluation harnesses.
• Experience with ML platforms / MLOps: model registry, CI/CD for ML, feature stores, monitoring stacks.
• Experience in health / regulated-data environments: security, auditability, traceability, careful UX for clinical contexts.
• Strong data unification experience: normalization, provenance, de-duplication, source-of-truth design.