About
Time to Power
Power-Capacity Intelligence for AI Infrastructure
What it is
Independent power and thermal engineering for AI data centers. We tell the owner of an existing data hall how much AI compute it can carry, which of thirteen electrical, thermal and physical constraints binds first, and what each step of extra density costs. We quote no equipment and take no margin on hardware.
Every number in a document carries an evidence class, from an assumption to an observation in operation, and the calibration record says how far the model has been checked against real sites. It is published, not hidden: evidence classes and calibration status.
Who is behind it
Time to Power is founder-led and pilot-stage. It is built by Vincenzo Grimaldi, a grid networks engineer working on the digitalization of high-voltage assets, with an M.Sc. from RWTH Aachen in cross-domain grid intelligence. His own site is igrimaldi.engineering. The engine underneath is GridForge Engine; its source is public at github.com/iceccarelli/GridForge-AI.
What we sell, and what we do not
- Fixed-fee engineering engagements and metered software, each priced before you start: engagements, pricing, machine interface.
- We sell no equipment, hold no inventory and take no margin on hardware. We name no make and no model.
- Engagement documents are produced and reviewed by a senior engineer, in screening mode. They are not an issued engineering opinion and carry no professional signature.
Where we are and how to reach us
- Based in
- Frankfurt, DE · Toronto, CA
- power@timetopower.ai
- Contact card
- contact.vcf, to add us to an address book
- Security
- disclosure policy and security.txt
For software and AI assistants
- llms.txt, an index of the site for language models
- openapi.json, the engine’s HTTP interface
- /api/cite, how to cite the constraint reference
- The thirteen constraints and the platform library