About Parisi Labs
Parisi Labs is building foundational world models for physical industry. We are developing models that learn how complex physical systems behave and reuse that understanding across forecasts, scenarios, and operational decisions.
Energy is our first proving ground. We combine historical and live data with operational context, bringing together machine learning research, data infrastructure, and software engineering to turn advances in modeling into useful technology for energy operators.
We are a small technical team working directly with the founders on our core models, systems, and products.
About the role
You will work directly with our co-founder and Chief Scientist, Matt Miller, to develop, implement, and test our core models. Matt's background spans MIT Media Lab, Bluefin Labs, nearly a decade in Twitter Cortex Applied Research, and leading Data and ML at Automattic.
We are looking for a researcher who can innovate, collaborate, and build foundation-scale models of complex, nonstandard systems. You will own independent lines of research while contributing to the team's central model-development work. This requires novel invention alongside meticulous rigor, strong engineering, and a commitment to real-world impact.
Our current work centers on generalizable time-series foundation models across electricity prices, demand, and generation on multiple grids. These systems reflect interactions among human behavior, energy markets, large-scale weather, and the physical grid. Open research questions include generalization, physical constraints and interactions, holistic world modeling, and simulation.
What you'll own
- Collaborate with the Chief Scientist to develop proprietary, state-of-the-art advances in time-series foundation models and world models.
- Train and test models on difficult, valuable, real-world forecasting tasks.
- Own independent research directions from hypothesis through implementation, experiments, ablations, and clear technical conclusions.
- Investigate multimodal inputs and event data, alongside methods for representing physical constraints and interactions.
- Work with engineers to productionize and ship successful research into systems that support real decisions.
- Extend advances beyond energy into domains such as supply chain, logistics, water, oil and gas, manufacturing, and predictive maintenance.
- Maintain reproducible experiments and communicate findings with the rigor needed to distinguish genuine progress from misleading results.
What we're looking for
- 5+ years of combined post-graduate and industry experience in ML research.
- A demonstrable history of research contributions and innovation that either genuinely advance the state of the art or deliver real business impact.
- Deep expertise in relevant technologies, such as time-series foundation models, world modeling, simulation, dynamical systems, and optimization.
- Creative problem-solving abilities and high adaptability.
- Strong coding and engineering skills: the ability to build in addition to invent.
- Relevant industry experience, such as energy, supply chain, or manufacturing, is a significant plus, not a requirement.
We are looking for PhD-level depth of knowledge and expertise, but recognize that this can be achieved in many ways; a PhD is not required. This is not a purely academic role. We need someone creative, passionate about impact, and able to execute and build while making full use of their technical expertise.
First 90 days
- 30 days: Reproduce the current model and evaluation system, understand its strengths and limitations, and identify a focused research opportunity with the Chief Scientist.
- 60 days: Own an experiment from hypothesis through implementation, evaluation, and failure analysis, with reproducible results and a clear technical readout.
- 90 days: Establish an independent research direction, deliver evidence-backed findings, and work with engineers to translate promsing results into the core system.
Why join
Research is central to the company's thesis. You will help develop genuine research intellectual property, advance the frontier of the field, and apply those advances to difficult, complex, valuable problems in physical industry. You will work directly with the founders and engineers who turn that research into useful technology.
Location and compensation
Location: Boston/Cambridge preferred; New York City can work for an exceptional candidate. This is a hybrid role - we expect in-person collaboration 3 days per week in person.
Salary range: $210,000–$275,000 USD.
Equity: meaningful early-company equity.