The company builds real-time multimodal models: fast, state-of-the-art voice models and the products that let people use them.
The founders met as PhD students at the Stanford AI Lab, where they invented State Space Models (SSMs), an architecture for training efficient, large-scale foundation models.
The company began as a research lab and now takes that work to customers through tools for creating, testing, and deploying voice.
The Voices team handles how customers find, clone, and design voices for their product in any language.
Nice-To-Haves are optional.
Responsibilities
Build product experiences for finding, cloning, and designing voices, from first listen through production use.
Design developer-facing APIs so every capability is as easy to use in code as in the product.
Turn early research into customer features such as voice cloning, localization, and voice design, working with the Research team to make new model capabilities into features such as multilingual voices and steering voices toward a target profile.
Build backend systems that train, store, and serve custom voices at scale, in the cloud and on-prem.
Raise the quality bar for each voice launch using internal tools, automated evaluations, and analytics on how voices sound and perform.
Work with Platform and Infra to serve every voice reliably at API traffic scale, with On-prem to make voices work across settings and hardware, and with customer-facing teams to help customers get a good voice on their own.
Requirements
At least 3+ years of experience building and running customer-facing software in production, ideally at a high-growth startup, including owning high-traffic systems. Shipped and maintained work at scale matters more than the exact number of years.
Strong product sense: you focus on customer outcomes, usability, quality, and developer experience, and you can argue for what to build and why.
Full-stack ability to own a product area across frontend, backend, APIs, data, and operations.
Fluent with coding agents and keeping up with the latest developments in the field.
Clear communication with technical and non-technical people, effective cross-functional work, and confident participation in product and design reviews.
Nice-To-Haves: experience building developer tools, creative tools, playgrounds, or API-adjacent products.
Nice-To-Haves: experience building AI-powered products and explaining model capabilities, limits, and quality trade-offs in terms users understand.
Nice-To-Haves: strong frontend skills with React and TypeScript, including polished, highly interactive workflows.
Nice-To-Haves: experience with audio, media, real-time systems, product analytics, experimentation, or evaluation infrastructure.
Nice-To-Haves: working knowledge of deploying, monitoring, and troubleshooting services on Kubernetes, especially across multiple clusters.
Conditions
The team works in person from offices in San Francisco, London, and Bangalore.
Visa sponsorship is offered case by case, but it depends on factors such as the role and work location, so it cannot always be guaranteed. A recruiter will discuss your visa needs from the first call.
The company ships quickly and does not trade away quality or design to do so.
The company is open and inclusive, and aims to give everyone the resources they need to succeed.
Equal opportunity: the company is committed to equal opportunity in hiring.
Benefits
Competitive base salary plus an attractive equity package (US employees only).
Fully covered medical insurance, with dental and vision, for you and your family (US employees only).
Parental leave of 9 weeks for paternity and 12 weeks for maternity (US employees only).
401(k) plan (US employees only).
Monthly commuter allowance to help you get to and from the office (US employees only).
Flexible PTO, taken as needed (US employees only).
Lunch, dinner, and snacks provided daily (US employees only).