Applied Researcher, Audio Post-Training

Our mission is to architect AI that learns from and interacts with the world like humans do.

We're pioneering the model architectures that will make this possible. Our founding team met as PhDs at the Stanford AI Lab, where we invented State Space Models or SSMs, a new primitive for training efficient, large-scale foundation models. Our team combines deep expertise in model innovation and systems engineering paired with a design-minded product engineering team to build and ship cutting edge models and experiences.

We're funded by leading investors at Index Ventures and Lightspeed Venture Partners, along with Factory, Conviction, A Star, General Catalyst, SV Angel, Databricks and others. We're fortunate to have the support of many amazing advisors, and 90+ angels across many industries, including the world's foremost experts in AI.

On the Audio Post-Training team, you’ll be building and improving the capabilities that define how the rest of the world interacts with our generative audio models. This team is where customer needs meet research, and covers the full spectrum of modeling from ideation through productionization. On any given day, you might design evaluations to reliably measure new capabilities, build processing pipelines to improve data quality, experiment with finetuning and reinforcement learning approaches to refine model behavior, and more.

This role is broad, and cross functional. Members of this team should combine broad research experience with a deep care about building to solve for customer needs and a strong sense of end-to-end ownership. You should be excited to synthesize customer complaints into a holistic understanding of capability gaps, to drive research efforts across data, model training, and evaluation to close those gaps, and to communicate those improvements to product and customer stakeholders. Ultimately, you will be responsible for creating the model experience that the rest of the world sees.

• Collaborate with product teams to understand and prioritize customer asks

• Cut through the ambiguity of vaguely described behavioral problems to make concrete research plans.

• Ideate and experiment across the full modeling stack, including data processing, synthetic data, SFT, RL, and evals to solve for high priority model capabilities

• Root cause failures in production models and understand how to fix them in future model iterations

• Decide which features and capabilities are ready for public launch

• Strong fundamentals in software engineering, machine learning, debugging complex systems, and the ability + desire to learn quickly.

• Experience building and ensuring quality of large multilingual datasets.

• Experience training and debugging generative models (speech, text, or multimodal), especially SFT, RL, synthetic data, and evaluation (both human and automated).

• Excitement about solving problems grounded in real customer needs, not just benchmarks.

• Bonus points if you have native proficiency in other languages!

   

Note: Cartesia participates in E-Verify and will provide the federal government with Form I-9 information to confirm employment eligibility after hire.

🏢 In-office policy: We’re an in-person team based out of offices in 🇺🇸 San Francisco, 🇬🇧 London and 🇮🇳 Bangalore. We love being in the office, hanging out together, and learning from each other every day.

Cartesia is an equal opportunity employer. We consider qualified applicants without regard to race, color, religion, sex, national origin, age, disability, veteran status, genetic information, or any other legally protected status.