Associate Director, AI/ML

Role Summary

Associate Director, AI/ML at Takeda's ShinrAI Center for AI/ML, based in Cambridge, MA. The role focuses on advancing the application of AI in pharmaceutical R&D, collaborating across data science, domain experts, and business units to drive decision-making and automation. The position emphasizes ethical development, cross-functional collaboration, and growth of AI/ML professionals within the organization.

Responsibilities

• Educate, demonstrate, guide, and enable the application of AI/ML in various pharmaceutical R&D operations and scientific domains
• Partner with data science teams, domain experts, and business units to identify and prioritize opportunities to leverage machine learning to drive decision making and automation across all levels of the R&D organization
• Translate business needs into clearly scoped machine learning projects, and take a hands-on approach to steer solution design and implementation
• Identify, monitor, and validate relevant external AI/ML developments, cultivate relationships with external domain experts and partners, and report and present emerging novel developments within the organization to further innovation and shape long-term strategy and governance
• Proactively build relationships across the company to inform your work and contribute to internal and external collaborations, through involvement in working groups, and the writing of insightful, engaging, and actionable opinion pieces that are easily digestible by internal decision makers and stakeholders
• Be the leading voice for building common capability and approaches and for adopting best practices
• Work in collaboration with our Ethics and Governance teams to ensure our AI/ML applications are developed ethically and provide broad benefits to our patients and business
• Help talented, driven, enthusiastic AI/ML engineers and data scientists across the company grow professionally
• Measure, document, and communicate impacts of the Center‚Äôs efforts

Qualifications

• A track record of partnering cross-functionally with a wide range of stakeholders and cross-functional teams to develop and deploy novel data solutions in production environments
• Demonstrated passion for making complex technology more accessible and the ability to communicate complex technical topics simply and convincingly to a wide range of audiences
• Demonstrated ability in translating big picture business and product ideas into micro use cases and has a strong focus on solving core problems to deliver simple solutions
• Experience recognizing and communicating the implications of emerging technologies
• Excellent communication, prioritization, and interpersonal skills, with a high level of attention to detail
• An advanced degree (M.S., PhD.) in mathematics, applied statistics, computer science, machine learning or similar. With 6+ years of experience architecting, building, launching, and maintaining end-to-end ML systems from whiteboard to production at scale across a range of models and platforms, such as experience in fine tuning large language models for domain-specific applications; experience in designing transfer learning strategy to enable learning from small datasets; demonstrated ability and authoritative knowledge in a variety of AI/ML problems and domains, with depth in at least two areas (computer vision, NLP, geometric deep learning, timeseries, reinforcement learning, multimodal learning, etc.); solid understanding of deep learning architectures and extensions; experience tuning, validating, optimizing, visualizing, and debugging these models in applied settings; familiarity with ML Ops concepts; experience in configuring and working in different coding environments and software engineering workflows
• An enthusiasm to ask questions and try and learn new things is essential
• Entrepreneurial experience is desirable
• Experience in life sciences and healthcare and experience in a complex global organization is a plus

Education

• Advanced degree (M.S., PhD) in mathematics, applied statistics, computer science, machine learning or similar