Remote
Deeter Analytics
Deeter Analytics is a new, well-funded startup building a new class of AI-driven trading system. We train foundation models on multimodal, petabyte-scale data sources, enabling innovation and discovery for intellectually curious specialists.
We believe that modern AI techniques developed at frontier labs will support the next generation of finance companies. Our goal is to create a fundamental trading model as capable as the most advanced AI systems, and deploy it on global markets.
You’ll work alongside a small group of elite engineers, researchers, and traders, in an environment defined by speed and ownership. We iterate quickly, with ideas moving from concept to experiment fast and the best ones pushed into production without bureaucracy
About the role
You will investigate a range of questions, like:
- How should long-horizon memory be represented for dynamic reasoning?
- What is the smallest capability preserving system, and where should compression happen: architecture, distillation, retrieval, memory, or runtime?
- How can we give the memory cell of our system a human chat interface?
- How do we safely increase model autonomy without losing observability, steerability, or control?
- When should a problem be solved by improving model, agent scaffold, context, or inference/runtime stack?
- How should we evaluate whether gains in offline reasoning survive contact with live deployment, latency, and market feedback?
And you will have responsibilities:
- Full development process from idea to prototype to production of architecture, data and infrastructure improvements.
- Design at the edge of current capability, from training on petabyte-scale datasets to deployment under tight production constraints.
What we’re Looking for
A rare researcher who combines first-principles thinking, production judgment, infra fluency, and founder-level agency - and who uses AI agents as a force multiplier.
Positive Indicators:
- Experience at frontier labs (or similar standard),
- Novel discovery through published work (NeurIPS/ICML/ICLR),
- Leading major open source project,
- Fully AI native workflows (maximal usage of codex, claude-code, cursor, etc),
- For Employees: A deployment of your research yielded >100x your salary,
- For Founders: Successful exit of a company you founded.