AI tools for your solar installation business — what works, what it costs, and how to start.

Yes, for a usable first draft. Aurora Solar and Enact pull the roof model from satellite imagery, run shading analysis from solar irradiance data for that address, and generate a system size, production estimate, and quote in under 30 minutes — no truck roll. The first draft is accurate enough to send to the homeowner and book the in-home site visit. The crew still confirms roof condition, electrical room space for the inverter, conduit run, and the main service panel rating before the contract is signed. The win is sending the homeowner a real quote the same day they fill out the form instead of three days later.

It gets you a solid design draft fast, but a few things stay on the crew. Aurora Solar and Enact read the roof from satellite or aerial imagery, lay out panels on the usable planes, run the shading model from the address, and spit out a system size and a yearly production estimate in kWh. That's enough to send a proposal and book the site visit. What the software can't see from space: the real shading once the trees leaf out, the roof's structural condition and whether it needs a re-roof first, the rafter spacing and attachment plan, and — the big one — the main service panel's busbar rating, which decides whether the array even fits without a panel upgrade or a line-side tap. A field tech confirms all of that before anybody signs. Treat the AI design as the draft that gets you in the door, not the final engineering. Find a local AI pro if you want the quoting tool dialed in to your typical roof types and utility territory.

Cash, loan, and PPA or lease — with at least two lenders on the loan side. Most residential solar sells on the monthly payment, not the total system price. The proposal tool should pull live rates from GoodLeap, Sunlight, Mosaic, Sungage, or Dividend (these five cover most residential solar lending in the US), show the qualifying monthly payment next to the cash option, and let the homeowner pick on the spot. Cash close rates are decent; financed close rates with a side-by-side comparison are markedly higher. One thing the software shouldn't do is promise a tax outcome: whether the homeowner can actually use the federal residential clean-energy credit depends on their tax situation, so the proposal should flag "check with your CPA" rather than bake a credit into the savings math as if it's guaranteed.

It handles the tracking, the document packages, and the reminders. The actual submission is still you and your office manager working the utility portal. Solar CRMs (Bodhi, Scoop, Solar Pro Tool) store per-utility application templates — PG&E, Duke, Dominion, FPL, Xcel, the rural co-ops — and tell you what each one needs before you start. Some tools can also read a homeowner's utility bill to pull the account number, rate schedule, and 12-month usage that the interconnection app and the proposal both need, so nobody's retyping it. They also fire reminders when the utility takes 6 weeks instead of 4 so you catch the silence before the homeowner does. Most utilities still need manual portal submission; a handful in larger states have direct API integration. Find a local AI pro to load your utility territories' templates and wire up the reminders.

Residential is the easier match. Aurora Solar, Enact, and OpenSolar all model residential rooftop accurately from satellite imagery and run on per-user pricing that a 1-5 crew shop can afford. Commercial PV — flat roofs, ground mount, carport, 200 kW and up — moves to enterprise platforms (Aurora Enterprise, Scoop, Solar Pro Tool) with custom pricing, and the design work is closer to engineering than a residential satellite pull. Most small installers run both: residential through Aurora at $199 a month, commercial on a per-project basis with an engineering partner.

It narrows the choice based on the roof model, the homeowner's utility rate, and the target system size. Aurora Solar and Enact compare panel options (REC, Q Cells, Silfab, Maxeon, the Tier 1 list) against the roof's usable area and shading map, then match inverter options (Enphase microinverters versus SolarEdge string-plus-optimizers versus a basic string inverter) to the layout. The AI does the math; the installer still picks the brand based on warranty terms, distributor relationship, and what's actually in the supply chain that week. Don't let the software pick on price alone.

Yes, and this is where most installers leave money on the table. Inverter manufacturers (Enphase, SolarEdge, SMA) all run monitoring portals that track per-panel or per-string production. AI monitoring tools (Bodhi, OptiSurge, the platforms built into the inverter portals themselves) alert you when a customer's PV system underperforms its expected output by more than 10-15 percent — usually a soiled panel, a tripped optimizer, or a bad connection. Catching the dip in week 8 instead of month 14 is the difference between a recall maintenance visit and a warranty fight. It's also the strongest moment to ask for a referral, because the homeowner just watched you solve a problem they didn't know they had.

That's exactly what post-install monitoring is for. The inverter portals (Enphase, SolarEdge, SMA) and the AI layers on top of them (Bodhi, OptiSurge) compare each panel's or string's actual output against what the design said it should be making for that weather and time of year. When one string drops 15 percent below its neighbors and stays there, that's a flag — a tripped optimizer, a soiled or shaded panel, a loose MC4 connection, or a failing module. The system emails you the alert so you schedule a service visit on your terms, instead of finding out 10 months later when the homeowner gets a true-up bill and a phone full of questions. The tech still climbs up and confirms the actual fault; the software just tells you which roof to go look at. Find a local AI pro to set the alert thresholds so you're not drowning in false alarms.

A residential solar lead typically closes 30-90 days after first contact, then takes another 60-120 days from contract signing to permission to operate (PTO). AI compresses three spots in that timeline: lead to first proposal (days to hours, via satellite quoting), proposal to signed contract (faster with financing options shown inside the proposal, plus follow-up cadence from the CRM), and signed contract to PTO (shaving 2-4 weeks off interconnection by catching utility delays early). The install itself — 1-3 days on the roof — does not get faster from AI. The paperwork around it does.