Job Description
This Google DeepMind product manager role sits in the Applied team in Mountain View, California, and it is one of the closest seats to the frontier you can get on the product side. Your job is to decide which Gemini capabilities reach Google's own products, in what order, and how anyone will know they are working once real people start using them.
The Applied team works across Gemini research groups and Google product areas. That makes you a translator. On one side are researchers pushing model quality forward. On the other are product teams who need those gains to feel useful, dependable and fast. If you have spent the last few years shipping software built on large language models and you want to move closer to where those models are made, this is the kind of opening that rarely comes around.
What You Will Be Doing
- Owning the roadmap for bringing Gemini capabilities into Google's first-party products. You will find high-value use cases, rank them honestly, and say no to the ones that sound exciting but won't move anything.
- Working side by side with research to decide what a new capability should do, how it gets benchmarked, and what it takes to turn a promising result into something reliable in production.
- Prototyping quickly with agentic harnesses. Instead of waiting on a polished spec, you will test ideas hands-on to show what a model can really do and to surface opportunities nobody planned for.
- Partnering with Google product teams to rethink, prioritize and build AI-powered experiences, and to give them a clear path to adopting Gemini for their most important use cases.
- Setting product strategy, roadmap and key features that make Gemini something both end users and developers choose to adopt.
Minimum Requirements
- A bachelor's degree, or practical experience that is equivalent
- At least 5 years working in product management or a closely related technical role
- Hands-on familiarity with advanced AI technology, especially large language models (LLMs)
Preferred Experience
- A history of building AI-powered consumer products
- Experience building machine learning (ML) products
Who Tends to Do Well Here
This is not a role for someone who only writes specs. People who thrive in an applied research-to-product seat usually share a few habits:
- They can hold a technical conversation with a researcher in the morning and a roadmap conversation with a product lead in the afternoon.
- They are comfortable with uncertainty, because what a model can do next quarter may look very different from what it can do today.
- They lean on evidence, such as benchmarks, experiments and user feedback, instead of gut feel.
- They write clearly. When several teams depend on your priorities, clear writing is what keeps everyone aligned.
Compensation Snapshot
The posted US salary range is $163,000 – $236,000 (USD), plus a 15% bonus target, equity and benefits. Google notes that individual pay depends on job-related skills, experience, and relevant education or training. Equity is granted at the discretion of Alphabet Inc., so it is best viewed as upside rather than a guarantee.
Tips for Applying
- Lead your resume with shipped LLM or ML products and the measurable outcomes they delivered.
- Show how you evaluated model quality, even if it was a simple internal benchmark.
- Highlight any work where you partnered closely with research or applied science teams.
- Mention hands-on experiments with agents or prototypes. It signals that you learn by building.
