Posted: 18 June
The role
Early stage frontier frontier AI lab founded by ex-OpenAI, MSL and Google DeepMind researchers.
Significant backing from tier 1 VCs. Currently in deep stealth mode
We're hiring a number of Research Scientists to work on conformal prediction and uncertainty quantification. You'll solve one of the hardest problems in applied AI...making LLMs and multi-agent systems reliably aware of their own uncertainty.
Your focus
- Conformal prediction for LLMs and agentic systems
- Uncertainty quantification under distribution shift and changing market regimes
- Selective prediction and abstention mechanisms
- Calibration and confidence aggregation across multiple agents
- Translating cutting-edge research into production systems
Experience we're interested in:
- Conformal prediction, conformal inference, uncertainty quantification and statistical learning theory
- Probabilistic modelling, online learning, or Bayesian methods
Ideal experience:
- Foundation models, generative AI, or multi-agent systems
- PyTorch, JAX, and large-scale ML systems
