Staff Product Manager, Agentic Experiences (Former Engineer)

Deepgram is bringing the world’s most natural interface, voice, to the technology we use every day. Our speech models and real-time infrastructure power AI systems that keep up with the way people actually speak. More than 5,000 organizations run conversational voice AI on Deepgram in production, including Twilio, Cloudflare, Sierra, Decagon, Vapi, Cresta, Granola, and Jack in the Box.

The Deepgram Voice Agent Platform brings together speech-to-text, text-to-speech and speech-to-speech, with voice-native foundation models available through cloud APIs or self-hosted and on-premises deployments. Built for accuracy, low latency and cost efficiency, Deepgram makes high-quality voice experiences practical to build and run at scale. More than 650,000 developers have used Deepgram to process more than 200,000 years of audio. Every day, more of the world’s conversations with AI run on our models.

Operating Principles:

Deepgram is a research company that productizes as fast as possible. Four Operating Principles shape how we think, decide, and work together:

1. Compress before you communicate. You think it through first, then lead with the bottom line.

2. Resist entropy. You question what already exists before adding more process or tools.

3. Solve the category. You fix recurring problems at the source, not one case at a time.

4. Work at the frontier. You go after problems no one has solved yet.

We look for people who already work this way, and these principles shape how we grow and evaluate our team.

We build AI that millions of people talk to, and we run the company the same way. AI is how we centralize work here, and is core to how we operate.

That means more than using the tools. We expect you to

• Build, not just use. Design your own workflows and agents, and use AI to remove whole categories of work instead of doing the same work faster.

• Own the output. AI can speed up the work. It doesn’t make anyone less accountable for it. You understand, verify, and stand behind everything you share, including what an agent produces for you.

• Spend the time you get back at the frontier. The capacity AI creates goes toward problems no one has solved yet.

• Keep pace. Models and tools change every month, and so will your work. You adopt what’s new quickly and drop what no longer earns its place.

We judge AI by what it helps you deliver and whether you stand behind the result, not by how often you use it. In interviews, expect to walk us through a workflow you built and where you drew the line between AI and your own judgment.

The Opportunity

The way developers find and adopt an API is changing. More and more, an AI coding agent discovers us, chooses the provider, writes the integration, and consumes our API, often with no human ever at the console. Deepgram is looking for a Staff Product Manager to own our product experience for that agent across its whole life with us: how an agent discovers and chooses Deepgram, how it integrates, how it uses the product in production, and how it verifies its own work. You will own that experience end to end, and you will build the system that keeps improving it as agent behavior changes. You report to the VP of Self-Serve.

This is a product management role in the conventional sense: you own the product, its direction, and its decisions, and engineering builds it. Two things set the role apart, and both are required — you are a former engineer who still builds to think and to prove a point, and you are deeply AI-native, with shipped work to show for it. You will prototype, read and write code, and reason with engineering at depth; your job is to own the product, not to be its implementing engineer.

What You'll Do

- Own the agent's experience of Deepgram across its lifecycle — discovery and recommendation, integration and onboarding, production use, and verification.

- Stand up a system that measures and optimizes every stage of the funnel for agents, and keep it current as agent behavior changes.

- Own the product surfaces specific to the agent experience: signup and authentication, the trial-key and token defaults and programmatic key provisioning, console onboarding, and the verification tooling that lets an agent confirm its integration is actually correct.

- Set the requirements for what the agent experience needs from the shared developer platforms — SDK ergonomics, the agent-readable documentation and llms.txt, the MCP server, the CLI, the skills package, and starter templates — and prototype the changes directly, in partnership with the team that owns those platforms.

- Stand up the operating system your work runs on — the rhythms of business, data-driven optimization, and the experimentation platform — by building it in-house or by researching and deploying the best tools available.

- Turn the scale of agent traffic into fast feedback loops, so the product improves as agents use it.

- Bring the product's point of view on agents as users: what they need, where they fail, and what to change, grounded in how models actually retrieve, choose, and integrate.

You'll Love This Role If You

- Were an engineer, moved to product to own outcomes, and never stopped building.

- Think like an architect and can design and stand up a self-optimizing system across discovery, onboarding, and integration.

- Have felt, first-hand, how an AI agent succeeds or fails at a real integration, and have strong opinions about why.

- Want to own a product that is becoming the front door of the business, at the moment it is becoming that.

- Are energized by being early — defining the practice, not inheriting it.

It's Important to Us That You Have

- Excellent product management judgment. You own product and roadmap, set direction, decide under uncertainty, ship outcomes, and lead cross-functional work without authority. You can show the results.

- A former engineer's depth (required). You were a senior software engineer, or more, before you moved to product. You architect and ship production systems, you read and write real code, and you reason with engineering at their level. You are not a vibe coder who assembles what a tool generates.

- Deep AI fluency, proven by shipped work (required). You have personally built and shipped AI software that goes well beyond prompt files and markdown — agents, MCP servers, CLI tools, agent and evaluation harnesses, real model-integrated tools — and it is public. Send us the GitHub; we will read the code, the commits, and the design.

- Proven ability to stand up a complete system from scratch — the rhythms of business, the reporting and optimization, the experimentation platform — yourself or in-house, or by researching and deploying the right tools.

- PLG and developer-product fluency. You understand how developers, and increasingly their agents, adopt APIs, and you understand product-led growth.

- The judgment to distrust a number or a passing test before you build on it. You ask whether it is real, as a reflex.

- Clear communication with executives: you lead with the decision, keep your method in reserve, and hold up under pushback without either caving or digging in.

It Would Be Great If You Had

- Built specifically for AI agents as the consumer — MCP servers, agent harnesses, CLI tools, agent-readable docs, tool definitions, or evals for agent output.

- Experience with voice, audio, or real-time streaming systems.

- A track record of open-source work with real adoption.

- Time in a company with both a self-serve and an enterprise motion.

Notice: We're aware of individuals impersonating Deepgram recruiters. All legitimate Deepgram recruiting communication comes from an @deepgram.com email address. If you've received a message claiming to be Deepgram, please forward it to careers@deepgram.com.