Ought is a product-driven machine learning lab building Elicit, an AI research assistant. Elicit uses language models to automate and support research processes like literature and evidence review. Elicit applies frontier technology to serious use cases, enabling our research team to understand in great detail where language models fail and how to mitigate such failures.
For more roles and to learn more about Ought, see the main careers page.
About the role
At Ought, "full-stack" means working in our Python backend (asyncio, mypy) and Next.js frontend, integrating various 3rd party and in-house ML APIs.
We have a hub in the Bay Area, but we are a remote-friendly company and are looking for excellent candidates no matter where they're based.
As a full-stack engineer at Ought, you will:
- Build intuitive interfaces that let researchers use language models as tools for thought
- Add new capabilities into our asynchronous Python app, composing results from internal and 3rd party ML APIs
- Contribute to discussions around product direction, technical architecture, and user feedback
- Add ways for users to quickly provide feedback to the tool to create an interactive training loop
- Act as custodian of our codebase, holding yourself and others to a high bar of code quality, testing, and resilience
Some engineering projects we're working on soon:
We think the person who will thrive in this role will demonstrate the following:
- Strong CS fundamentals and software engineering background
- Hands-on experience with automated testing for complex software systems
- Experience with, or ability to quickly adapt to, a typed Python web app
- Ability to quickly convert ideas into tests and prototypes
- Clear communication about engineering topics to a diverse audience
You'll especially enjoy working at Ought if you have:
- Interest in language models (training, finetuning, evaluation) and natural language processing more generally
- A startup mindset. We expect to measure our impact in part by the people whose lives we improve through better reasoning and models of the future