For the last few years, a core group of procurement experts, solar and storage optimization engineers, and software programmers have been actively developing new technology that’s flipping the script on the traditionally slow and complex process of investigating, vetting, and buying solar modules and energy storage equipment.
To understand the journey from home-grown development optimization tool to robust web-enabled marketplace, I offer my perspective of the operational and development issues we faced as Anza’s technology was born.
Since my early days, I’ve been a white space problem solver, eager to bring form and function to amorphous business problems. I enjoyed the thrill of writing code that solved a particular problem, yet I wanted something more; something that made a clear business impact. That’s a big reason why I went back to school to get my MBA – to learn how to be a leader who could make more tangible impacts.
Throughout my professional career, I’ve had a chance to see the benefits of those impactful moments, of connecting the thrill of coding with successfully managing a team to deliver meaningful results. Before I joined Borrego, I’d worked as chief information officer and chief technology officer in technology, and non-technology, companies. In the role at Borrego, I was attracted to the opportunity to do something different and consequential with my skill set. Borrego (and the solar industry) had reached an inflection point — exponential changes in the size and scale of solar projects was ushering in a new period of rapid growth for Borrego, known internally as the “Gigawatt Revolution” to coincide with the the shift from megawatts to gigawatts. I was excited to be in on the ground floor opportunity and help change the way the company leveraged technology across its businesses.
A couple years prior to developing the Anza engine prototype, we had early conversations about how we could bring together and automate the enormous amount of data points, processes, and internal IP used by the in-house engineering and project teams across the full solar project lifecycle. Borrego CEO Mike Hall coined its first name, “the solar easy button” — visualized as a metaphorical red button that could quickly rank modules using our proprietary project data and module optimization processes.
What emerged from our initial work was a customer-friendly way to provide procurement as a service, leveraging our project design automation data to quickly find and demonstrate the best value modules for any project based on net present value (NPV). We knew from years of developing projects that the cheapest module was not always the best choice for long term project returns — and this new easy button could quickly show the relative value each module option would bring to any project.
While we moved toward developing and launching the Anza platform, the solar industry faced seismic shifts in module availability. Even before COVID, we had begun to see IPPs and developers taking on module procurement to safe harbor modules before the ITC declined. But increasing volatility due to the pandemic, supply chain disruptions, tariffs and other policies complicated procurement, and it’s become increasingly more common for developers and IPPs to take on the procurement when pricing and supply chain difficulty has made it too risky for EPCs to do so.
Since IPPs are not usually module procurement experts, we knew our new platform could simplify, optimize, and bring some degree of confidence to what could often be a time consuming and complex approach to module purchasing, even in the best of times. With Anza’s NPV ranking tool, we were able to improve an outdated, analog process and give large-scale module buyers something they’ve never had before — a simple yet deeply analytical way to quickly procure the best modules and level up the value of their projects.
No origin story would be complete without some discussion of the major challenges going from concept to reality. In Anza’s case, the number-one challenge we had to overcome was to create a way to efficiently capture and sustain the technical and domain-specific knowledge that’s core to our optimization tool and NPV ranking table. Trying to get programmers and software designers to interact with and understand the sophisticated nuances of solar module pricing and product specifications was challenging. We spent a year trying to build a parallel tool for code optimization for the in-house software development team to work with, before coming up with a different solution.
The answer to this challenge comes from our coding and engineering roots. We thought, “what if we could teach solar engineers to code?” To facilitate this, we had to come up with a particular microservices-based architecture. We taught our solar engineers enough Python to write deterministic pure functions that directly contributed their domain expertise to the product. This allows for authentic, highly specialized, and continuously improving calculations that drive the NPV ranking tables — without the bottleneck and translation required to hand those critical details to a software developer. This solution also freed up the software team to focus on the technical scaffolding and interface, without having to learn the details of solar & storage optimization. We’ve found that this approach is one area that differentiates us.
I’m excited by not only what we’ve accomplished so far with Anza and the gigawatts of traction the platform has already achieved, but also by the vast potential for expanding its capabilities. As we fine-tune the optimization engine for both modules and energy storage components, Anza will become even more automated, with additional layers of analytics and functionality baked in for the user. We also believe we can apply this transformative technology to other products or markets.