Edviro: AI operator for energy equipment (YC Summer 2026), Skoor 40
What they built
Edviro describes its product as "AI That Operates Energy Infrastructure." [source]
The company says it connects to meters, HVAC systems, batteries, solar and other energy equipment. [source]
It claims to detect and diagnose problems, forecast demand, simulate fixes, create work orders, coordinate maintenance and control equipment where authorized. [source]
It says its "energy world models" learn each site's behavior, choose actions and verify results against operational, meter and billing data. [source]
It is based in San Francisco, CA, and is tagged Deep Learning, B2B and Energy. [source]
Record
Measured signals (n = 4, Skoor 40, confidence 0.8):
- YC batch Summer 2026, weight 15. [source]
- YC company listing, weight 10. [source]
- Hiring per the YC directory, weight 10. [source]
- Team of 2, weight 5. [source]
The status is listed as Active and the stage as Early. [source]
All four signals come from the YC directory, observed 2026-10-14. [source]
There are no code repositories, organization records or specific job listings in the evidence.
Why now
Not in the evidence.
Inference: The pitch depends on software being able to act on equipment, not just report on it. Whether the timing is right depends on customer willingness to authorize control. The evidence does not test this.
Unknowns
- Founders, their backgrounds and technical depth: not in the evidence.
- Customers, sites connected, pilots or revenue: not in the evidence.
- Funding beyond YC: not in the evidence.
- Which roles are open: not in the evidence. The listing shows hiring, but no job records were captured. [source]
- Whether any site runs with equipment control enabled: not in the evidence.
Inference: The Skoor comes entirely from YC directory status. It measures selection, not traction or the team.
Questions for the room
- You claim to control equipment "where authorized." How many sites have granted control today? What equipment types are they, and who signed off?
- You say results are verified against meter and billing data. Show one site's before-and-after numbers. How did you separate your effect from weather and occupancy?
- "Energy world models learn how each site behaves." How long does onboarding a new site take before the model picks actions? How much of that is per-site manual work for a team of 2?
What would move the Skoor
These would raise it:
These would lower it: