MANY MINDEDAn agent that haggles down your solar quote before you sign
A startup called AgentSolar AI is aiming its tools at one of the least examined costs in residential solar: the sales markup. Founder Tursun Ablekim, a former researcher at the National Renewable Energy Laboratory and First Solar, says the selling process itself can add 30% to 40% in soft costs to a system's final price. The company was written up on July 18, 2026.
The product is a set of agents. A Quote Evaluation Agent reads proposal PDFs of up to 40 pages, pulls out the financial line items, and checks them against a market database. A Company Research Agent cross-checks the installer's history and reviews. If the homeowner decides to proceed, a third agent runs a structured negotiation, generating counteroffers and sending emails, then notifies the homeowner when a fair contract is ready. Every automated message, the company says, is reviewed by a human administrator first.
The hardware side of solar has been getting cheaper for years; the soft costs have not. Ablekim puts the scale of the gap at roughly $2 billion in US homeowner overpayment in 2025, drawn from over a million data points, and $4,000 to $6,000 of avoidable acquisition cost, from door-knocking, lead buying, and high-pressure sales, on a single 8 kW system. If a cheap agent can do the price discovery and haggling an informed buyer rarely has time for, the markup that inflates rooftop solar gets squeezed from the demand side.
Ablekim predicts that within five years AI could handle the whole home-energy journey, from roof modeling to virtual-power-plant management. Worth watching, but worth keeping in proportion: the overpayment estimate, the soft-cost percentages, and the acquisition-cost range are all his own framing, and the five-year outlook is a founder's forecast, not a demonstrated result.
Source: pv magazine