Bali Solar Energy - lead-to-proposal system

A lead-to-proposal system that turns cold solar enquiries into structured opportunities, preliminary proposals, and qualified sales conversations, with minimal manual work from the team.

See the live platform

---

Project snapshot

 

 

The situation

Bali Solar Energy is a new solar company operating in Indonesia. It needed a sales process that could handle a growing number of potential customers while keeping the amount of manual work required from its team under control.

Solar sales carry a particular challenge. A customer usually starts with a simple question: "How much would solar cost for my villa?" But giving a useful answer requires information about the property, electricity consumption, location, meter capacity, and roof. Traditionally, a salesperson has to collect all of that by hand before even a first estimate can be prepared.

That creates real friction. Response times are slow. Salespeople spend their time gathering basic information instead of working with qualified opportunities. Manual calculations and proposals can carry mistakes. And as the number of leads grows, simply adding more manual work does not produce a sales operation that scales.

The goal was clear: build a sales system that can take a cold lead from first contact to a qualified solar opportunity with minimal manual work.

 

The challenge

The system had to work with people who might know very little about solar. A cold lead should not have to understand panel configurations or technical requirements before getting a useful answer. So instead of asking for engineering detail, the system asks for a small amount of information any property owner can provide.

The first estimate needs four things:

1. Average PLN bill: typical monthly electricity spend.
2. Location pin: a Google Maps location for the property.
3. PLN meter capacity: for example 2,200 VA, 5,500 VA, or 7,700 VA.
4. Roof type: such as concrete, tile, metal, or alang-alang, with a photo where available.

The same process also had to work no matter where the lead came from. A customer might start through WhatsApp, Telegram, or the website, while the workflow behind it stays connected. So the real task was bigger than building a chatbot. It was building the system around the sales process.

 

Decisions before code

The most important decision was to make the initial qualification simple enough for a cold lead to complete without any human help. Instead of a long form or an interview with a salesperson, the system asks for the four pieces of information that feed the first calculation, through whichever channel the customer already uses: WhatsApp, Telegram, or the website. The company meets people where they already communicate instead of forcing every lead into one channel.

We also decided the first proposal should be generated automatically. The point is not to replace the engineering team. It is to move the lead much further through the sales process before a person needs to step in.

From the four inputs, the system prepares a preliminary proposal that includes the recommended system size, an approximate number of panels, an estimated price, estimated monthly and annual savings, the payback period, estimated electricity generation, an installation roadmap, and the next steps. The customer receives it by email within minutes.

The proposal is preliminary, not a final engineering specification. Site inspection, engineering validation, and final pricing stay human responsibilities. That boundary was intentional, and it is the same principle throughout: automate the preparation, keep the professional decisions with professionals.

 

How the system works

The whole process reduces to four stages.

Capture. A customer starts through WhatsApp, Telegram, or the website, and the system collects the four inputs. No salesperson has to gather the basics first.

Estimate. AI turns those inputs into a preliminary proposal, moving the customer from "how much does solar cost?" to "here is what a system for my property could look like." The lead receives a structured proposal, not just an answer in a chat.

Qualify. At the same time, the collected information lands in the CRM. Instead of a bare "I want solar for my villa," the sales team receives a structured opportunity with the information already gathered and the preliminary assessment attached, then offers the customer a free audit as the next step.

Follow up. Not every lead replies immediately, so the system can follow up automatically, helping the team stay in contact with leads that do not respond to the first message. The process keeps working after the initial conversation instead of depending on someone remembering to chase manually.

 

The build

The system was built in six weeks across three two-week sprints, starting on 22 June 2026, by a team of one CTO, two developers, and one designer. techlead.fit acted as hands-on CTO across product decisions, architecture, and delivery.

The customer sees one simple process. The complexity sits behind it.

 

Release and validation

The MVP launched after the six-week build and is now live. This is an early version. We are actively evaluating the process, measuring how leads behave, and adjusting the workflow based on what we learn.

At this stage the objective is not to claim a finished, optimized sales machine. It is to establish the infrastructure and validate the process in real conditions. We are monitoring the full journey, from lead through qualification, estimate, proposal, CRM, audit, follow-up, and into the sales conversation, and the system is built to be adjusted as real lead data shows where it can improve.

 

Outcome

The immediate outcome is a live MVP that turns an incoming solar enquiry into a structured sales opportunity with minimal manual work.

Bali Solar Energy, live. Screenshot from the public product. No customer data shown.

A typical journey now runs:

Cold lead ↓ WhatsApp / Telegram / Website ↓ Four pieces of information ↓ AI-powered assessment ↓ Preliminary solar proposal ↓ Structured CRM record ↓ Free audit offer ↓ Automated follow-up ↓ Qualified opportunity for the sales team

The system is still being validated, so we are not publishing conversion or revenue numbers yet. The more important result at this stage is that the company now has an operational foundation that can process substantially more opportunities without the same increase in manual sales work. Because the workflow is built around a shared backend, CRM, and automation layer, it can be extended as the company grows into other Indonesian regions.

The same ecosystem also includes a content factory that automates a large part of the company's SEO and social content operation, from location-specific articles and social posts to images, videos, scheduling, and keyword-driven content. This creates a second acquisition layer around the sales system: content generates visibility, visibility generates leads, and the same backend moves those leads into qualification and sales.

 

Why this is a techlead.fit case study

The starting point was never "let's build an AI chatbot for a solar company." It was a business problem: how do we process more solar opportunities without making the sales team handle every early-stage lead by hand?

The answer required more than AI. It needed customer-facing interfaces, lead qualification, property information, estimation, proposal generation, CRM, follow-ups, and analytics working as one system. AI handles the repetitive analysis and preparation. The sales team handles the decisions that need human judgment. The engineering team handles the final technical validation. And the business gets a process it can measure, improve, and scale.

Capture the lead. Collect the right information. Automate the preparation. Keep professional decisions human. Measure the process. Scale what works.

---

Looking to automate your sales process?

If your team spends too much time qualifying leads, preparing initial proposals, and following up by hand, the first question is not which AI tool to add.

It is: which parts of the sales process can be automated, which decisions still need people, and how should the whole system work together?

That is the kind of system we build at techlead.fit.

Relevant services: Technical Audit · CTO Sprint · MVP Build