
1. Technology overview
Godela is a company that builds AI models for the physical world, founded by former engineers from Apple, Google and Intel who turned Stanford research into a product. It targets the weeks or months engineers spend on simulations, data preprocessing and parameter tuning to answer a single design question.
The platform trains a physics-constrained model on a customer's existing simulation data, CAD files or experimental results, then lets engineers ask what-if questions and optimise a design in seconds. The site shows applications in data center thermals, electronics cooling, aerodynamics, electromagnetics and structures, and offers a pilot to prospective customers.
2. Highlighted features & specifications
- Train: builds a physics-constrained model from the customer's simulation data.
- Explore: answers what-if questions on airflow, geometry and materials in natural language.
- Optimize: searches thousands of configurations to find a better design.
- Domain coverage: data center thermals, electronics cooling, aerodynamics, electromagnetics and structures.
- Custom models: turns any physical system into a queryable model from a question, geometry or scenario.
3. Potential application
A data center operator in Vietnam could test rack airflow layouts before installation. An electronics manufacturer could evaluate heat sinks for a new device. A mechanical design team could screen structural variants without full finite-element runs.
4. Benefits
Engineers get simulation-quality answers in seconds rather than days. Existing simulation data is reused instead of discarded. Teams can explore many configurations and rank them before building hardware. The site's example shows a data center cooling case where peak temperature fell from 72.1 C to 64.8 C.
5. Market trends & opportunities
Engineering teams are adding AI surrogates to speed up costly simulation workflows. Demand is strong in data center cooling and electronics design. In Vietnam and Southeast Asia, electronics manufacturers and data center builders face the same simulation bottlenecks.
6. Other information
Y Combinator Spring 2025 company (YC X25). Founders are former engineers at Apple, Google and Intel with Stanford research roots. The website offers a pilot request.
One platform connecting startups, corporations, research institutions, regions and investors











Playbook & Handbook · February 6, 2026
Winning Accelerator Playbook 2026 by BambuUP
RETAIL E-COMMERCE: AI AND REAL-TIME DATA ARE TRANSFORMING HOW RETAIL BUSINESSES OPERATE








































