Role Summary
Design, build, and fine-tune AI systems that power real healthcare impact. Success looks like shipping reliable models to production, improving accuracy/latency/cost, and running reproducible experiments. This role sits within the core AI team and partners with product, data, and engineering to turn research into scalable products.
Responsibilities
Design, train, fine-tune, and evaluate models (LLMs, embeddings) for production use cases.
Build reproducible data/ETL pipelines and experiment tracking.
Implement MLOps best practices (versioning, CI/CD for models, monitoring, rollback).
Run benchmarking, A/B tests, error analysis; document findings and decisions.
Create internal tools/SDKs for inference, evaluation, and prompt engineering.
Deploy via REST APIs and optimize for performance and cost.
Communicate results to technical and non-technical stakeholders.
Qualifications
Skills: Python (essential); TensorFlow/PyTorch; Hugging Face; Pandas/NumPy/Scikit-learn; ETL; visualization (Matplotlib/Plotly/Tableau/Power BI); NLP & GenAI (LLM fine-tuning, embeddings, prompt engineering); statistics (hypothesis testing, regression).
Nice to have: C++/Java; publications or patents; Unreal Engine exposure.
Environment: Solid Linux background for development and deployment.
Education/Experience: Relevant degree or equivalent practical experience.