
Founded by Vivek Hazari, Daniel DiPietro, Ziray Hao & John McCambridge
Vivek, Dan, John, and Ziray: the team behind Lumenary. They met each other during their first few weeks at Dartmouth and have been best friends ever since. They have lived together, worked together, traveled the world together, and have even summited some notable peaks together (don’t ask about Mount Shasta).
The team brings together engineering from Palantir, Jane Street, IMC Trading, and Flatiron Health. They have built enterprise-grade AI products, a capital allocation system that controls billions, a data engine that processes petabytes–they all build and deploy software quickly in contexts where it cannot fail.
Solar began as an internal tool that they used to ship enterprise demos faster than any tool on the market. They decided to productize and focus on Solar completely.

As a team of four engineers who’ve used nearly every AI-powered app builder. Projects started off great, but they found themselves in an endless loop of “revert, and try again.” Most of their time was spent waiting and hoping as the LLM made changes one-at-a-time.
Even when the app seemed to work, it was unclear what were real features versus frontend demoware. They lacked an understanding of the architecture, dataflow, and what vulnerabilities were left open. It felt like these platforms had prioritized making neat frontends and bolted on real functionality as an afterthought.

The team loved three things:
So they combined those to make Solar.
Solar’s AI agents collaborate alongside you on a multiplayer canvas. You can see the relationships and data flow between database tables, backend Python logic, and frontend pages.

Solar makes apps production-worthy by combining LLM-generated code with deterministic code generation for databases, services, and authentication. This makes your apps maintainable and deployable past the MVP, and able to integrate nicely with existing infrastructure.
They are the most agentic platform on the market: Solar will read logs, interact with your application, and work for minutes or even hours unattended to get it perfect.