Work with us at futuresearch

Evaluate and improve state-of-the-art forecasting and agentic LLM web research. Start with an open problem, devise an innovative solution, and build the evals that prove it works.

A senior role with wide scope. In a typical month you might ship a feature in the FutureSearch app, extend the pipelines behind our benchmarks, and work out why an agent pool times out under load.

Support the development of tasks for training and benchmarking frontier models. For careful researchers who can get to the bottom of subtle information retrieval problems.

Quality-check the evaluation tasks we develop for training and benchmarking frontier models. For meticulous investigators who can establish whether a hard research question really has the answer we shipped with it.

Some of our most consequential work comes from people whose job title doesn't fit that of engineers and researchers. We are looking for people with a proven track record and a keen interest in AI to join us in building AI that predicts the future.