Member of Technical Staff

We're a team of machine learning engineers training task-specific generative models for psychology. Our goal is to build an AI therapist to help people change their mind and their lives in the ways that they want to. We partner with organizations around the globe and power use cases, including AI-assisted crisis text response, while securing best-in-class datasets to power our models.

Success to us means every human being in need of support having somewhere to go. We're a well-funded, seed-stage startup backed by top-tier tech investors involved in Huggingface, ElevenLabs, Replit, Captions, Shopify, Plaid, Notion, Canva, Twitch, Airtable, and others

We're building a powerful team by empowering our engineers with the autonomy, flexibility, and resources to do their best work. We dream big and iterate fast. If that sounds like home, we'd love to hear from you.

The Role

As Member of Technical Staff, you'll work with our founding team across ML research and product development. To be successful in this role, you have to care about how these models perform to help people in the real world - more than how they perform on artificial evals. We ship a lot. You'll be able to work at a faster pace than almost anywhere else while writing high-quality code and producing meaningful scientific insights.

You may work on data collection, curation, continued pre-training, ablation studies, evals, supervising the creation of hand-crafted data, preference optimization, and state-of-the-art reinforcement learning research. You'll also contribute to our end-user product, improving user experience through your work on our models and model orchestration.

You'll be working on the latest language models, ranging from open-source 70B / 405B to frontier models through our deep partnerships with the largest AI labs. You'll read papers and identify state-of-the-art techniques for us to learn from; and contribute to our core ML research : aligning language models towards successful long-term trajectories.

Our application backend is written in Kotlin and our ML stack (PyTorch) utilizes modern tooling in the ML space, including some that we've developed in-house in Typescript. We write high-quality, typed, Zen code.

About you :

Note : We expect our MTS to be highly autonomous and intelligent individuals who may not fit our exact mold. Feel free to apply even if you don't exactly have the right experience, if you're ready to learn quickly on the job.

• 5 years developing deep learning models in PyTorch, TensorFlow or JAX, include 3 years in a production environment.
• Experience fine-tuning language models, like Llama.
• Experience with software engineering on a product, e.g. React, Swift, Kotlin, Java, etc., including a strong understanding of modern software architectures.
• You're fast-paced and pragmatic. You'd rather prove out an idea through quick MVP code than present a slide deck to explain it.
• You can explain complex ideas to non-technical people
• You understand why deep learning is magic.

What We Offer