Why Solar Companies Need an AI Second Brain (and How to Query Your Margins in Plain English)

Most solar founders have tried AI. You open ChatGPT, paste an angry client email, ask for a description of a hybrid inverter, and get something usable.

Then Monday arrives. Three supplier updates land on your phone. An installer asks where the warranty certificate is for a system commissioned six months ago. The technical office asks which inverter gives the best margin on flat roofs. You open ChatGPT again, remember it knows nothing about your supplier discounts, your stock or your local council requirements, and close the tab.

When the chat window closes, the AI stops working for you. You are back to being the manual router between WhatsApp, email, the ERP and your team.

Solar companies do not need another generic chatbot. They need an AI second brain: a structured base that gives AI access to how the business actually runs.

Useful AI Power = Model × Context × Processes × Tools

Most people think it all depends on the AI model. It matters, but it is not enough: for AI to do real work in an installation business, four things have to work together.

$$\text{Useful AI Power} = \text{Model} \times \text{Context} \times \text{Processes} \times \text{Tools}$$

1. Model ($M$): the AI you rent (ChatGPT, Claude, Gemini or an open model). It changes every few months, and your competitors can use the same one.
2. Context ($C$): what the AI needs to know about your company. Component price lists, preferred brands (Huawei, Fronius, SMA), labour rates, regional subsidy rules, municipal permit requirements.
3. Processes ($P$): how a task gets done. What inputs are needed, what steps to follow, and where a human has to sign off.
4. Tools ($H$, from the Spanish Herramientas): the connections that let AI act. Email, your ERP (STEL Order or Odoo), cloud storage, digital delivery notes.

If any factor is zero, the result is zero, even with the best model. You rent the model; context, processes and tools are yours.

• Context without tools produces good answers you still have to copy and paste.
• Tools without context produce generic or wrong documents that damage your credibility.
• Context, processes and tools together turn repetitive admin into workflows that run on their own.

What a solar second brain looks like

It is not another enterprise suite with months of onboarding. It is a clean set of open files (Markdown notes in Obsidian or private cloud storage) that acts as the single source of truth for the business.

                  ┌────────────────────────────────────────┐
                  │       COMPANY SECOND BRAIN             │
                  ├────────────────────────────────────────┤
                  │ 1. Context: Tariffs, Equipment, Specs  │
                  │ 2. Processes: Quoting, Permits, Audits │
                  │ 3. Tools: ERP, Email, Delivery Notes   │
                  └───────────────────┬────────────────────┘
                                      │
           ┌──────────────────────────┼──────────────────────────┐
           ▼                          ▼                          ▼
   Instant Answering         Automated Proposals       Natural Language Queries
"Which warranty applies?"   "Draft quote in 30 mins"    "Net margin by project?"

In a working installation company, it usually has three folders:

1. Context

• Equipment specs: verified attributes for your standard panels, inverters and batteries.
• Supplier agreements: discount levels and contacts for distributors like BayWa r.e. and Krannich Solar.
• Municipal rules: permit requirements, ICIO tax rebate percentages and prior declaration formats for the towns where you install.
• Company voice: how you address clients, standard warranty clauses, payment terms.

2. Processes (SOPs)

• Site survey procedure: required photos, electrical measurements, cable paths.
• Quoting procedure: how to size strings, calculate payback, format proposals.
• Permit checklist: required annexes and engineer sign-off, as described in our guide on automating technical memories and building permits with AI.

3. Tools

• Connections to your operational software: syncing project cards in the ERP, attaching photos from mobile delivery notes, drafting email replies with real context.

Querying the business in plain language

Once context is organized and connected to operational data, you stop spending weekends cross-referencing five spreadsheets to answer one question.

You can ask:

• “What was our average net margin on residential versus commercial installations last month?”
• “Which permit applications have been pending for more than three weeks?”
• “Show me clients who installed more than two years ago and have not had a firmware check or a maintenance offer.”

Because the AI has your actual numbers, catalog costs and project statuses, it cross-references them immediately and gives you an answer with the figures behind it.

Team autonomy: when office staff and crew leads can reach the same knowledge base, interruptions drop. The team checks the second brain for manufacturer procedures or municipal rules instead of interrupting the founder every twenty minutes, which is the classic founder bottleneck.

What it looked like at Vivim Solar

We worked with Vivim Solar, a ten-person company in Sant Cugat del Vallès near Barcelona. Its founder, Xavier Castellvi, was the single point of contact for every decision, technical question and quote confirmation.

Building the company second brain:

• captured knowledge that lived only in the founder’s and head technician’s heads;
• structured equipment references, catalog prices and quote workflows in the team’s working language;
• gave them a base to build their first automations on during our two-month program.

The result was 5 to 10 hours a week of administrative load removed, faster quoting, and a business that runs with more order and less firefighting.

Build on what keeps its value

The AI landscape moves fast. New models and tools appear every month, and chasing each one is a distraction for a busy owner.

A company second brain holds what lasts:

• tools and models will keep changing;
• your context, verified processes and operational data are assets you own;
• when a better model shows up, you plug it into the existing second brain and get the benefit immediately.

If you want to build this in your company and compare notes with five non-competing solar founders, see our AI training for solar installers.