Applied GenAI Intern: Build Practical AI Agents

Genentech; San Francisco (CA); On-site; USD 100,000 - 120,000; Full time

Benefits offered by this job

Stipend provided

Job summary

Genentech is seeking a 2027 Spring Intern to explore GenAI applications within the CMG domain. You will build a working prototype of programmable GenAI agents, evaluating how instructions, skills, context, and tool access affect performance.

The project will leverage Claude Code and agent concepts to demonstrate reusable approaches for business use cases. Join a 6-month, full-time paid internship based in South San Francisco, with a stipend and collaboration across data scientists, business

Qualifications

• Foundational understanding of ML and generative AI, including large language models.
• Experience using Python to conduct experiments, analyze data, or prototype AI applications.
• Ability to design experiments, critically evaluate results, and communicate findings clearly.
• Interest in applying GenAI to practical business problems and collaborating with domain experts.

Responsibilities

• Define the use case and evaluation approach for a CMG business need.
• Design and prototype specialized GenAI agents with tasks, skills, data and tools.
• Assess and de-risk model components, compare hosted vs open-weight options.
• Conduct systematic experiments on configuration, context, memory and tool selection.
• Evaluate business value, reliability, and cost against a baseline.
• Develop a reusable capability and share findings with stakeholders.
• Bring latest techniques from academia into the organization.

Skills

Foundational ML knowledge

Python experimentation

Experiment design & evaluation

GenAI/LLM concepts

Education

Master's Degree (enrolled)

Master's Degree (earned)

PhD (enrolled)

Tools

Claude Code

Open-weight models

Encoder-based classifiers

Job description

Genentech is seeking a 2027 Spring Intern to explore GenAI applications within the CMG domain. You will build a working prototype of programmable GenAI agents, evaluating how instructions, skills, context, and tool access affect performance.

The project will leverage Claude Code and agent concepts to demonstrate reusable approaches for business use cases. Join a 6-month, full-time paid internship based in South San Francisco, with a stipend and collaboration across data scientists, business