For a decade, while the hardware costs for rooftop solar have plummeted, the soft costs — permitting, customer acquisition, and design — have remained stubbornly high. The industry is a fragmented landscape of smaller installers using a patchwork of legacy technology stacks. The result is 100-day project timelines, delays in getting homeowners a quote, and high customer churn, even as demand continues to rise.
One small installer argues that this is fundamentally a software problem, and that solving it is key to taking solar mainstream. And it’s overhauling its entire business model to pursue that bet.
Monalee, a direct-to-consumer installer founded in 2022, is shedding its capital-intensive construction company business model and replacing it with that of a high-margin tech provider. Now, armed with a fresh round of venture capital funding, it’s aiming to do for solar what Shopify did for e-commerce, or what Carvana did for car sales. Artemis, the rebranded company centered on Monalee’s technology, aims to be the engine that powers everyone else’s rooftop solar sales, CEO and co-founder Walid Halty told Latitude Media in an exclusive interview about the move. “The vision is ultimately, if you want to buy solar and storage…you enter an address, you can buy it, and we can install it the same day. That’s the ultimate goal.”
Walid, a Tesla solar sales alum, previously founded Colossus, a solar marketplace that was acquired in 2023 by the technology platform PX. There, he saw customer satisfaction crater as soon as project leads were handed over to installers that got bogged down in manual design and permitting processes. At the time, Colossus opted to stay in its own lane. “We were like, surely someone is going to fix all of this other stuff, let’s stay focused on what we’re doing,” Halty recalled. “Five years later, no one did anything about it.”
Monalee initially attempted to fill that gap by building a digital buying experience for rooftop solar, giving customers an instant price and system design without ever talking to a salesperson. As the installer, Monalee was able to cut out sales commission costs, making its projects cheaper for homeowners.
But when they set out to scale Monalee’s construction work nationally, Halty and his cofounders found that in a post-25D tax credit environment it would require an immense amount of capital, right at a time when capital for asset-heavy climate startups was increasingly scarce.
Instead, the team decided to cut its capital-heavy installation business loose and go all in on software late last year, licensing its internal platform to installers as their front-end design and sales engine. The company has now sold its remaining installation assets to a network of regional partners who now handle the trucks and ladders, Halty explained — though it retains its existing software relationships under the new Artemis name.
Today, Artemis’ tech stack uses computer vision and machine learning, Halty said, training on more than 200 million data points from real-world installations, and roughly 1.9 million human annotations. It generates system designs and production estimates for solar and batteries in seconds, and will eventually also include roofing and HVAC. If the AI runs into an edge case, like a tricky roofline or an odd setback rule, Artemis’ team steps in to correct the design.
Key to making those estimates underwritable, Halty explained, is Artemis’ approach to the permitting process: The company is developing agents capable of identifying specific jurisdictional requirements, auto-filling permit templates, and managing the email follow-ups with local building departments. The result, according to Halty, is a more than 50% reduction in software-related soft costs, quoting times compressed from a week to a few minutes — and in the case of one of Artemis’ largest customers, seven-figure annual savings.
Returning to the software playbook
Artemis isn’t the first to apply automation to the soft costs problem. Aurora Solar, a software company founded in 2013, uses LiDar and advanced 3D modeling to create site assessments. However, while Aurora has become the industry standard for certain tasks, it is largely a backend tool, targeting the inner workflow of solar installers. Artemis, meanwhile, is targeting the consumer-side with a “Shopify-style” approach: an embeddable self-serve widget installers can add to their website.
Larger companies have also attempted a soft costs software fix, but haven’t been particularly successful. After acquiring SolarCity in 2016, Tesla attempted to commoditize solar through a “buy it now” button that aimed to strip away the complexity of a traditional solar sales cycle, leveraging satellite imagery and mapping data to estimate a home’s solar potential. The company quickly hit logistical hurdles in the form of permitting delays, site-specific roof complexities, and the costly overhead of maintaining national field crews. Tesla eventually scaled back its solar ambitions — that is until earlier this year when the company announced it was leaning back into rooftop solar, manufacturing its own mount and module.
Artemis, Halty said, is avoiding those challenges by decoupling the software rails from the physical installation — which would in theory allow it to scale at software rather than infrastructure speed, and potentially even generate venture-scale returns. Earlier this year, Artemis closed a $6 million funding round, designed to help the company pivot away from its capital-heavy Monalee legacy and toward a pure software play. The round, which follows the roughly $13 million the company raised as Monalee, will underwrite more investment in Artemis’ AI design engine, Halty said, as well as expanded engineering and product teams.
For Copec Wind Ventures, the San Francisco-based VC arm of the Chilean energy conglomerate that led the round with Long Journey Ventures, Artemis represents an attempt to bring the latest tech developments in Silicon Valley to an industry that hasn’t historically operated on the “move fast and break things” timeline of tech companies.
In a landscape where AI is becoming ubiquitous, it’s Artemis’ data sets that make it stand out, explained investor Mark Dryden. Because Artemis is selling its software to the wider industry, rather than applying it to its own installation work, there’s always the risk of imitation and competition, Dryden acknowledged.
“But AI is as strong as the data underneath it, and Artemis has done a fantastic job of capturing all of these licenses to drone data sets and everything else,” he said, pointing to the more than 250,000 drone data collection flights and 6,000 registered jurisdictions on the platform to date.
While that data is U.S. specific, Copec sees Artemis’ tech as an opportunity to bridge the gap between Silicon Valley’s AI advancements and the burgeoning rooftop solar market in Latin America, where there’s even less standardization and digitization. Work to prove that Artemis’ machine learning models are transferable across markets is already underway in Chile, where Artemis is integrating with Copec’s energy and e-mobility subsidiary Flux.
That early deployment, already live, has included swapping in local aerial imagery providers, tuning the models to Chilean irradiance and utility conditions, and validating designs against real projects in Flux’s pipeline. According to Dryden, that work shows that the Artemis design engine can adapt to different energy patterns and lower-quality imagery without needing the decade of training it had in the U.S. And Chile is just the jumping-off point for Artemis in Latin America.
That global scalability is central to Halty’s “Shopify for solar” thesis. If Artemis can enable a regional installer in Santiago to operate with the technical sophistication of a national giant in California, in theory the same playbook could be replicated across dozens of rooftop markets worldwide — without Artemis ever owning trucks or warehouses in those countries.
