Calda AI
Ben Wandelt, PhD
Co-founder
Bloomberg Distinguished Professor at the Johns Hopkins AI Institute. Founding director of the Lagrange Institute, Paris. Awards include the Gruber and Bessel Prizes. Previously Caltech, Princeton, IAS and the Sorbonne.
Accelerating engineering.
AI for Physics
We build high-fidelity AI surrogates for physics simulations, making them up to a million times faster than existing solvers. By resolving the simulation bottleneck — from plasmas, to fluids, to materials — we are on a mission to revolutionise engineering and unlock transformative new industries.
01 Our approach
New speed, new possibilities.
Imagine
Engineering design reimagined.
Fast, differentiable simulation enables testing of thousands to millions of design options in silico before investing in a single prototype, providing both human engineers and agentic systems with foundational new tools for design. By combining fast AI surrogates for physical simulation, sophisticated search algorithms and agentic workflows, we are bringing in a new era of AI-automated engineering design and discovery.
Optimize
Bringing high-fidelity physics to real-time.
Fast simulation allows engineers to monitor and predict what is happening in their processes, in detail and in real time. This opens up new possibilities for both controlling complex processes online, and designing completely new manufacturing-processes and strategies offline. These new capabilities are transforming both how we make stuff, but also what it is possible to make, opening up entirely new industries — from nuclear fusion to materials discovery and quantum compute hardware, and beyond.
Understand
Physical data analysis without compromise.
By drastically reducing the cycle-time for simulation, we’re also able to make more accurate predictions and gain a better understanding of the world. Calda makes it possible to simulate detailed small-scale physics and effects that were previously too expensive, enabling high-fidelity predictions that were impossible before. By building data fusion and analysis pipelines with fast and accurate simulators embedded inside them, Calda enables you to infer what happened in your process or experiment and why, now in as much detail as is physically possible given your data.
02 Applications
Complex physics unlocks transformative industries.
Plasma
Fusion & plasma physics
Fusion promises abundant clean energy, but unlocking it requires understanding and controlling turbulent plasma at extremely high temperatures and densities. A single detailed reactor simulation can take weeks on a supercomputer, putting critical design decisions behind a formidable simulation bottleneck.
Calda’s plasma surrogates deliver accurate predictions up to a million times faster, across diverse physical regimes covering particle-based (WarpX), gyrokinetic (GKW and GX) and magnetohydrodynamic (MHD) solvers. We are opening the door to faster reactor design, deeper physical insight and new approaches to plasma control, bringing down the timelines for limitless clean energy from nuclear fusion.
Explore plasma with us
Fluids
Fluid dynamics
Mastering fluids is critical across many emerging technologies, from harnessing clean energy from the wind and ocean, to more efficient transport, and next-generation cooling systems. Yet resolving complex flows is so computationally demanding that engineers must limit the designs they explore and the physics they include.
Calda builds fluid surrogates that combine speed and accuracy with long, stable rollouts and generalisation to unseen geometries—making detailed flow predictions practical across more designs and operating conditions than previously imaginable.
Explore fluids with us
Materials
Material deformation
Lighter structures, more durable components and new manufacturing processes depend on understanding how materials deform—and how that shapes their performance. Detailed finite-element simulations can provide that insight, but their cost puts it beyond everyday engineering decisions.
Calda’s material deformation surrogates run 500,000× faster than industry-leading solvers, enabling predictions of mechanical properties inside your product in real-time, where no measurement device can reach.
Explore materials with us03 Our team
Meet the founding team.
Niall Jeffrey, PhD
Co-founder
Faculty at King's College London, AI Institute and Physics. Award-winning researcher in physics and AI. Previously Imperial, UCL and École Normale Supérieure (Paris).
Justin Alsing, PhD
Co-founder
Former faculty at Stockholm University. Pioneer in simulation-based inference and AI for accelerating physical simulation. Previously Imperial, Oxford, Simons Foundation.
Lucas Makinen, PhD
Co-founder
Physics-AI Fellow at the University of Cambridge. Physics PhD and expert in AI for physics and statistics. Trained at Princeton, Harvard, the Sorbonne and Imperial College London.
Mona Wilcke, PhD
President
Serial entrepreneur and CEO of Stockholm University Ventures with over 20 years experience building and scaling research-based companies.
Hamza Bokhari, PhD
Commercial
Deep tech business development executive: fundraising, licensing and partnering across Europe, the US and Asia.
04 Start a conversation
Waiting days for simulations?
Tell us about the physics you simulate and the possibilities you want to explore.
Contact us info@calda.ai