Overview As a senior AI/ML researcher, you will define the long-horizon ML vision for a healthcare-focused organization, guiding the development of a domain-specific foundation model and a high-reliability inference system. You will shape how frontier models are leveraged, decide when to build proprietary domain models, and sequence capabilities into monetizable, customer-facing features while adhering to safety and regulatory constraints. You'll serve as a senior technical advisor on frontier-model integration, data strategy, evaluation, and safety architecture, partnering with product, engineering, clinical, and compliance teams to ensure safe, reliable, and differentiating AI capability over time.
Responsibilities- Architect and guide a long-horizon ML strategy for healthcare foundation models
- Define technical vision for frontier-model use and proprietary domain models
- Oversee data strategy, evaluation harness design, and safety architecture
- Lead cross-functional collaboration with product, engineering, clinical, and compliance teams
- Ensure AI systems are safe, reliable, economically viable, and capable of long-term differentiation
- Advise engineering and science teams on ML systems engineering, training, and deployment
- Drive high-stakes, regulated-domain ML shipping and governance practices
- Balance research depth with pragmatic product sense to build enduring capability
Key requirements- 10+ years in AI domains
- 5+ years of language/multimodal model tuning and optimization
- Proven track record shipping large-scale AI/ML products
- Experience training or adapting large-scale models (LLMs, multimodal, MoE)
- Strong grounding in distributed training, RLHF/DPO, retrieval and knowledge integration
- Experience with evaluation harness design and ML systems engineering
- Familiarity with clinical data, workflow constraints, and safety/regulatory practices (ISO-aligned or equivalent)
- cross-functional collaboration
- storytelling and technical leadership
- pragmatic product sense
- large-scale model training/adaptation
- RLHF/DPO
- retrieval and knowledge integration
