Senior Data Scientist, Gen AI Application

We're passionate about delivering on Our Promise to improve the lives of patients and create healthier communities for all. We foster a culture of inclusivity, integrity and creativity while boldly pursuing answers to the world's most complex health challenges and transforming society.

Genentech's Data, Digital, and Analytics (DDA) team is dedicated to solving complex healthcare challenges and improving patient outcomes. DDA empowers business partners across Commercial, Medical, and Government Affairs (CMG) to make impactful decisions by leveraging data, analytics, and AI/ML to enable fast, targeted actions in rapidly evolving business contexts. DDA fosters a unified understanding of customers, actions, and outcomes by transforming the business insight supply chain from the traditional reactive service model to a modern proactive product model, which integrates analytics and insights seamlessly into CMG's evolving digital, data, and automation platforms, creating scalable solutions and eliminating silos. In DDA, you will work as a trusted, objective advisor and expert, recommending critical decisions and actions to be taken with credibility and a focus on driving measurable impact. You will be part of a diverse, inclusive team that reflects the world we serve, thriving in a welcoming culture built on collaboration and innovation.

The Opportunity

The Sr. Data Scientist - Generative AI Application Specialist builds what users actually touch: the conversational, agentic, and GenAI-workflow applications that turn our AI capabilities into products people trust and rely on. This is a front-end-led, full-stack role for the age of AI. You will own the experience layer where reasoning agents, streaming responses, and generative interfaces meet real users, and you will partner closely with infrastructure and product teams to

make those experiences fast, legible, and reliable. We are not looking for a traditional data scientist, a UI-only developer, or someone who only "vibe-codes" with an assistant. We are looking for a strong engineer who is redefining what front-end means when the interface is a living, reasoning system - while representing the team's expertise across the broader Genentech and global Roche organizations.

• Build AI-Native Applications: Design and ship production front-ends for conversational and agentic experiences - chat, multimodal, and generative/adaptive UI - that make complex GenAI workflows feel simple, transparent, and trustworthy.
• Master Real-Time & Streaming Interaction: Architect the real-time layer
• (WebSockets/SSE, token streaming, live agent status, interruptibility, optimistic and latency-aware UI) so conversations and long-running agent tasks feel responsive and alive.
• Design the UI as an Agent Surface: Build interfaces that agents can read and act on - tool-call visualization, human-in-the-loop controls, and "UI-as-context" patterns that let assistants ground their answers in what the user is actually seeing.
• Deliver Conversational & Multimodal UX: Craft chat and (over time) voice/multimodal interactions, surfacing memory, state, and reasoning in ways that build user trust and keep humans in control.
• Own Front-End Craft & Engineering Fundamentals: Deliver well-architected, accessible, performant React/TypeScript applications backed by a real component and design system. This is solid, durable front-end engineering - not throwaway prototypes or assistant-generated code you cannot stand behind.
• Bridge Front-End and Agentic Systems: Partner with infrastructure and orchestration teams to design the contracts between the app and the agents/GenAI workflows behind it, shaping how state, memory, and context flow across the full stack.
• Set Direction & Elevate AI Literacy: Act as a technical coach and pattern-setter,
• influencing technical priorities and elevating how the wider organization thinks about interaction, trust, and transparency in AI-native products.
• Operate with Rigor in a Regulated Environment: Manage high-stakes internal and external partnerships while ensuring all work adheres to global compliance, security, and regulatory standards.

• Bachelor's degree and 5+ years of experience building software, with a strong emphasis on front-end and application engineering.
• Front-End Engineering Depth: Expert-level React and TypeScript (or an equivalent modern framework), with deep command of component architecture, state management, accessibility, and performance - a real engineer who writes and reviews production code, not solely AI-generated output.
• Real-Time & Streaming Experience: Hands-on experience with WebSockets, SSE, and streaming interfaces (e.g., token-by-token responses, live status, and interruptible or long-running tasks).
• GenAI Application Fluency: 3+ years building GenAI-powered or conversational
• products, with practical understanding of agents, prompting, tool-calling, RAG, and the trade-offs of GenAI workflow design.
• Conversational & Generative UI: Experience designing chat, multimodal, and/or
• adaptive/generative interfaces, and a thoughtful approach to keeping humans in the loop, building trust, and grounding responses in what is on screen.
• API & Systems Integration: Comfort integrating front-ends with agent/LLM backends and APIs (REST, event-driven patterns, MCP/A2A), and reasoning about cost, latency, reliability, and observability across the AI system end to end.
• Communication & Collaboration: Ability to partner with product, design, and
• infrastructure teams and to translate complex technical trade-offs for non-technical stakeholders.

Preferred

• AI-Native Interaction Patterns: A point of view on how front-end is evolving in the age of AI - generative UI, agent-readable interfaces, UI-as-context, and novel interaction models beyond the chatbox.
• Design & Craft: Strong design sensibility and fluency with design-to-code workflows (Figma), motion/polish, and building and maintaining a design system.
• Voice & Multimodal: Experience with voice and other multimodal GenAI interfaces.
• Full-Stack Range: Comfort reaching into the agent and orchestration layer (LangChain, LangGraph, AWS Agentcore) when the experience requires it.
• Industry Experience: Experience in healthcare, pharmaceutical, or other highly regulated industries.
• Security Protocols: Familiarity with common security protocols and building software under regulatory constraints.
• Enterprise Navigation: Proven ability to influence technical priorities and manage partnerships within a large-scale, global corporate structure.
• Resilience & Risk Management: Experience leading "fail-forward" initiatives where rapid prototyping and business-value trade-offs are prioritized.