Foundation models for scientific and industrial data.
Expert AI: foundation models for the data that runs science and industry
We define Expert AI as foundation models built for scientific and industrial data: continuous and irregular time signals, spatiotemporal fields, multimodal scientific recordings, time-stamped event logs from real organizations. The kind of data that doesn’t look like text, doesn’t fit transformer assumptions, and rarely leaves the institutions that hold it. This isn’t a fine-tuning problem. It requires architectures designed for the data, infrastructure designed for how the data lives, and a tech stack of AI components purpose-built to assemble foundation models for new domains.
General-purpose AI was never built for science and industry. The data is sparse, noisy, distributed, and governed by strict rules. It demands expert intelligence: models that understand the structure of a domain, not just the surface of its data.
Our approach is co-foundation with domain experts: AI and domain science evolving together. We build domain-specialised foundation models trained to master a field’s data, constraints, and logic from the ground up. We work hand in hand with the scientists, clinicians, and engineers who know the domain best. Our models learn from distributed sources through decentralised architectures, keeping sensitive data where it belongs.
We anchor in brain and neuroimaging, the most complex signal of all and the hardest class of expert data we know how to characterize. In parallel, our cross-domain transfer programme works on extending the same components and methods to open new application domains.
We invest in demonstrators that turn research into tangible proof, working with partners on concrete use cases that test our models in real conditions and inform the next iteration of the science.
Our team finished 2nd out of 1,183 in the NeurIPS 2025 EEG Challenge, and we publish at Tier-1 AI conferencesincluding NeurIPS, ICML, and CVPR. This dual focus, rigorous science and applied impact, defines our approach.
Sigma nova is looking for interns (starting in Q1 / early Q2 2027)
You will work alongside researchers on the practical challenges of training and deploying models for complex neural time series. The project will be shaped around your interests and experience, with a focus on building something measurable and usable: a benchmark, a reusable tool, or a working prototype.
1) Efficient models and deployment
Explore how to make deep learning models smaller and faster while preserving their capabilities. Possible directions include model compression, distillation, quantization, and deployment on resource-constrained devices, with opportunities to build a real-time demonstration.
2) Scalable processing of brain signals
Develop and evaluate ways to represent and process large, diverse brain-signal datasets more efficiently. You could work on reducing computational costs, improving data pipelines, or making models easier to use across different recording configurations.
3) Reliable experimentation and benchmarking
Build tools that help researchers compare models and understand practical trade-offs between performance, speed, and resource usage. This may involve reproducible evaluation pipelines, profiling, or integrating research prototypes into a robust codebase.
These are alternative directions, not a checklist: together, we will define a focused project for the internship.
Strong Python skills and experience with PyTorch
Solid deep learning fundamentals and an interest in efficient ML systems
Enjoyment of writing clean code, debugging, and measuring what actually improves
Experience with model deployment, signal processing, or performance optimization is a plus
Prior neuroscience or EEG experience is welcome, but not required
A research-driven startup at the interface of AI and neuroscience
Close mentorship and collaboration with researchers and engineers
Access to large brain-signal datasets and high-end GPU resources
Room to shape the project and contribute to tools used by the team
Opportunities to contribute to a publication, depending on the project and results
Prescreen recruiter
Technical screen with a Research Engineer
Onsite interviews (coding / research discussion)