Enterprise Case Study · Digital Transformation
End-to-End AI-Powered Automation for a Premium Beauty Salon — Cali, Colombia
01 · Executive Summary
Glam Color Studio operated a 1,400+ client portfolio managed entirely through manual processes within the Litzo ERP platform — with zero digital customer engagement, no automated communication layer, and an undifferentiated web presence that failed to reflect the brand's premium positioning.
This engagement delivered a full-stack digital transformation: a branded marketing website, an AI-powered WhatsApp conversational agent, a custom CRM dashboard seeded with real client data, an automated Litzo data pipeline, birthday campaign automation, and a production-ready deployment architecture — all engineered within a single intensive development sprint.
02 · Business Challenge
| Operational Challenge | Business Impact |
|---|---|
| Client records stored exclusively in Litzo ERP with no external access layer | Zero ability to trigger automated outreach or segment client base |
| No digital communication channel between appointments | High client churn risk; no re-engagement mechanism for inactive clients |
| Static or absent web presence | Brand visibility gap; premium positioning undermined by digital absence |
| Manual birthday and follow-up campaigns | Missed revenue opportunities across 839+ clients with confirmed birthdays |
| No conversion layer on Meta ad spend | 15 conversations generated from paid campaigns with zero confirmed bookings |
| No structured stylist data capture post-service | Recurring service history unavailable for personalized recommendations |
03 · Solution Overview
The solution was architected as a cohesive seven-component ecosystem, where each layer feeds data and context into the next, eliminating manual intervention across the full client lifecycle — from initial discovery to post-service re-engagement.
04 · Data Intelligence Layer
A full forensic analysis was conducted on the Litzo Excel export, revealing critical data quality gaps that informed both the migration strategy and the segmentation logic for automated campaigns.
05 · AI Conversational Agent
The WhatsApp agent was engineered with a dual-track architecture that dynamically routes clients based on CRM recognition — delivering a personalized, contextually-aware experience to returning clients while efficiently qualifying new prospects.
| Step | Bot Action | Data Captured |
|---|---|---|
| 1. Identification | Phone number checked against Supabase CRM | New vs. returning client flag |
| 2a. New Client | Warm welcome greeting + service inquiry | Service category of interest |
| 2b. Returning Client | Personalized greeting with name + visit history | Confirmation of intent |
| 3. Service Qualification | Inquiry into desired procedure | Specific service requested |
| 4. Chemical History | Single consolidated prior-process question (if color) | Last chemical treatment date/type |
| 5. Photo Request | Request back-of-head photo + reference inspiration | Visual assessment data |
| 6. Human Handoff | Graceful escalation to human advisor | Full context passed to stylist |
06 · Technology Stack
The architecture was deliberately selected to maximize operational capability while maintaining optimized cost structures — leveraging serverless and edge deployment paradigms that eliminate traditional infrastructure overhead entirely.
| Layer | Technology | Role | Cost Model |
|---|---|---|---|
| Frontend | HTML/CSS/JS + Cloudflare Pages | Marketing site + CRM UI | Zero-infrastructure overhead |
| AI Engine | Claude Haiku API (Anthropic) | WhatsApp bot intelligence | Pay-per-token, optimized margins |
| Bot Server | Node.js + Express + Railway | Webhook handler + API orchestrator | Cost-efficient serverless runtime |
| Database | Supabase (PostgreSQL) | Client CRM + campaign data | High-throughput serverless DB |
| Automation | Make (Integromat) | Data pipeline middleware | Event-driven, zero-compute overhead |
| Scraping | Puppeteer + Node-cron | Litzo ERP data extraction | Self-hosted scheduled extraction |
| Image Engine | Python + Pillow | Birthday personalization | Zero marginal cost per unit |
| Messaging | Meta WhatsApp Business API | Outbound + inbound messaging | Per-conversation pricing model |
07 · Outcomes & Roadmap
All core deliverables were reviewed and approved by the client during a live presentation session. Diego confirmed approval of the website design and system architecture, commissioning a Phase 2 expansion immediately.
"The measure of a digital transformation is not the technology deployed — it is the degree to which every client interaction becomes effortless, personal, and intelligently orchestrated."
Glam Color Studio Digital Transformation · Portfolio Case Study