From zero to a system in 3 weeks: the Sando Capital case

Sando Capital is a financial intermediary in Mallorca. Javier Sandoval founded it without a digital presence: no website, no acquisition system and no customer database. Leads arrived through WhatsApp and referrals — when they arrived at all.

Every conversation with a potential investor required a manual 20–30-minute call to understand their profile, capital, time horizon and risk tolerance. Only then did Javier know whether booking an appointment made sense. Customer knowledge lived in his head and in scattered messages.

The problem was not the product. It was the architecture

Javier knew how to do his job: connect investors with suitable investment products. What he lacked was a system. A system that acquired leads, qualified them, stored them and allowed him to work with data rather than chats.

This is the difference between a business that scales and one that does not. When customer knowledge lives in WhatsApp, every new lead has a marginal cost of 30 minutes. When it lives in a system, every new lead takes 30 seconds to read a record.

What we built in 3 weeks

1. A custom website

A professional website presenting Sando Capital’s services and generating leads from the first contact. Designed for first-time investors: clear, quiet, with a single CTA — book an appointment. It is not a generic landing page. It is the entrance to the system.

2. Scheduling with Cal.com

Integration with Cal.com so investors choose a time slot directly. Javier’s calendar synchronises automatically. No double scheduling, no “does Tuesday suit you?” messages, no context lost between channels.

3. An AI qualification agent

Before confirming the appointment, an AI agent talks with the investor. It asks about capital, time horizon, previous experience and goals. It analyses the answers and suggests the right product — or rules out the lead if they are not a good fit.

Javier receives an analysed investor record before the call. He does not waste 30 minutes on discovery — he arrives knowing what to offer. This changes the equation: the AI agent does not replace Javier; it eliminates the work Javier should not be doing.

4. Database + CRM

Every lead arriving through Cal.com or the website form is stored in a centralised database. Javier has a dashboard showing all contacts, their status (new, qualified, booked, closed), the suggested product and interaction history. Customer knowledge no longer lives in WhatsApp. It lives in the system.

Why 3 weeks, not 3 months

We build on Foundation, our monorepo template: Nuxt 4 + NestJS 11 + PostgreSQL 17 + Redis + BullMQ + LangChain. We do not start from zero — we start from a system. Authentication, the task queue, the ORM, Cal.com integrations and the AI layer already exist as modules. They are assembled and adapted.

This is not speed for its own sake. It is speed because we do not repeat work already done. The difference between 3 weeks and 3 months is not efficiency — it is not rebuilding what already exists.

Results

• 0 → 1: From nothing to a website + complete system.
• −70% of the time spent qualifying leads: The AI agent does the work before the call.
• 100% of leads in the database: Nothing is lost in WhatsApp or emails.

The website they made for Sando Capital impressed me. Then they integrated Cal.com with AI so the system itself reviews investors’ needs and suggests the right product. It works as a perfect first filter.

— Javier Sandoval Haro, Founder of Sando Capital

Architecture is what enables AI

Sando Capital’s AI agent works because it has architecture underneath. Without a database, Cal.com or a website, the agent would have nowhere to read or write. You cannot plug AI into chaos. First organise, then digitise, then deploy intelligence.

This is the process we follow in every project: audit before building, build before adding intelligence. To see how to apply it to your business, book a free 30-minute discovery call.