361 STATION - B2B SaaS, built solo end-to-end

I designed and shipped a multi-tenant CRM with an AI agent. Alone. Brand, copy, product and deploy directing AI agents from concept to production.

Client
361 Station (own product)
Category
Product & AI
Year
2026
Impact
SaaS live in production - app.3meia1.com.br. AI serving real customers 24/7 on WhatsApp. Stack: Next.js · Supabase · Vercel · LLMs · Meta Business Suite. Built by 1 person directing AI agents.

The platform, in production

361 STATION is a conversational CRM for service businesses - clinics, law firms, consultancies. This is the real panel, live: the AI suggests the Focus of the Day, tracks automatic conversations and bookings, and shows what the team saved - hours, human-free conversations and avoided cost.

Real 361 STATION dashboard: overview with AI-suggested Focus of the Day, conversations, bookings and savings

This is not a concept case study: it's a SaaS in production, with real customers being served by AI on WhatsApp right now. One person - directing AI agents - designed the brand, wrote the copy, shaped the product and shipped the deploy. What used to take a whole team, 361 STATION delivers as a platform: 24/7 service, automatic qualification, a full calendar and operating costs in free fall.

100% Responsive

The whole operation fits in a pocket: the owner tracks conversations, pipeline and bookings from anywhere. The AI works; the panel shows it.

361 STATION on mobile: the full panel, responsive, in the palm of the hand

Liz, actually working

A real conversation: Liz - 361's AI agent - replies on WhatsApp, offers the professional's open slots, books, and sends the Google Meet invite. When the customer asks for a human, the AI pauses and escalates automatically; the team takes over in one click, with full context and captured data on the side.

Real conversation in 361 STATION: AI agent Liz offers time slots, books and sends the Google Meet invite; a human takes over in one click

Visual pipeline, no refresh

Every conversation becomes a card that moves through the funnel on its own as the AI qualifies and books. True realtime: the team watches the pipeline update live, with the assigned agent and source channel on every contact.

361 STATION kanban sales pipeline with New Contact, Qualifying, Booked, Active Client and Lost stages

Every agent, custom-trained

I designed the Master Panel where each AI agent is configured: operational behavior, interaction limits, business hours, and the rules for when to call a human - complaints, complex questions, price negotiation. The client's onboarding becomes the AI's training.

Master Panel configuring agent Liz: behavior, limits, timezone, and the escalation-rules wizard

End-to-end white-label

True multi-tenancy: each client gets a signup link, fills in their business details, and the answers feed the agent's creation. Their own logo, colors and domain - the end customer never knows a third party is involved.

361 STATION client onboarding: white-label signup link generator and the platform login screen

Brand, copy and sales site

Beyond the product, I designed the brand and wrote all the copy - "Your team serves more, without working more". On the sales site, the demo shows Liz replying in real time to real clinic, law and aesthetics conversations.

Website demo section: "Watch Liz chat like a real person", with a real conversation in a phone mockup

Concept to deploy. Solo.

I architected the product and shipped it to production directing AI agents through the build - Next.js on Vercel, Supabase backend, realtime. Role: solo - brand, product, copy and deploy. Live at app.3meia1.com.br.

361 STATION interface details: contact modal, kanban card and the Ask AI field