Tribo · Theoretical Case Study · 2025–2026

AI Automation Suite for an
Aesthetic Medicine Clinic

Theoretical Documentation — Business context, solution design, technology rationale, and measurable outcomes for a full-stack automation engagement.

9AI Bots Live
73Services Scoped
70%CRM Integration
24/7Lead Capture
Document Contents

Table of Contents

01
Executive Summary
The business context and high-level outcomes of the engagement.
02
Business Challenge
The operational inefficiencies that necessitated the automation suite.
03
Solution Overview
Strategic approach and the four pillars of work delivered.
04
Fibot AI Chatbot Suite
Architecture, conversation design, and deployment of the bot suite.
05
CRM & Data Strategy
Custom CRM system, patient data management, and Lizto integration.
06
Technical Infrastructure
The full technology stack and zero-infrastructure-overhead architecture.
07
Meta Ads Integration
How paid media campaigns connect to the automation ecosystem.
08
Results & Business Impact
Measurable outcomes delivered across all project dimensions.
09
About Tribo
Firm profile and service offering for international clients.

01 · Executive Summary

Transforming a Manual Intake Process
into an Autonomous Revenue Engine

An aesthetic medicine clinic in Cali, Colombia, operating across nine active service verticals — facial harmonization, aesthetic dermatology, and non-surgical cosmetic procedures — engaged Tribo to architect and deliver a full-stack commercial automation suite.

The mandate: eliminate manual patient intake, create a scalable lead qualification engine tied to Meta Ads campaigns, and consolidate fragmented patient data into a unified, actionable system. The engagement delivered a production-grade AI-powered conversational bot suite (Fibot), a bespoke CRM, a corporate digital presence, and a bidirectional Lizto CRM integration pipeline.

Scope of Engagement
Full-Stack Automation Delivery

AI chatbot suite · Custom CRM · Corporate website · Meta Ads integration · Lizto data pipeline · Analytics dashboard

Client Profile
Aesthetic Medicine Clinic

Cali, Colombia · 9+ service lines · Facial harmonization, aesthetic dermatology, non-surgical procedures · Active Meta Ads per vertical


02 · Business Challenge

The Operational Inefficiencies
We Were Hired to Solve

Fragmented Lead Capture at Scale

The clinic ran independent Meta Ads campaigns for each of its 9+ service verticals. Each campaign generated inquiries through generic channels with no automated pre-screening, no clinical history collection, and no intelligent routing to the appropriate specialist advisor.

Response Latency
Prospect Attrition After-Hours

Lead response was dependent on staff availability. Prospects engaging outside office hours received no acknowledgment — causing direct loss of Meta Ads ROI at the point of first contact.

No Pre-Screening
Unqualified Advisor Inquiries

Advisors received unqualified inquiries with no clinical history, no reference photography, no service context — making every initial consultation operationally expensive.

Disconnected Data
Siloed Patient Ecosystem

Patient data existed across Lizto CRM, WhatsApp threads, and manual records. No unified patient profile, no cross-system journey view, no campaign attribution model.


03 · Solution Overview

Strategic Approach & Scope of Work

Tribo designed and delivered a modular, scalable automation suite organized across four interconnected pillars — each addressing a specific operational gap identified in the discovery phase.

PILLAR 01
AI Conversational Lead Capture

Nine AI-powered chatbots (Fibot), one per service vertical, deployed to Cloudflare Pages. Each bot intercepts Meta Ads traffic, conducts structured clinical pre-screening, collects reference photography, and delivers a qualified lead to the specialist advisor via WhatsApp. Built on a deterministic state-machine architecture with Groq LLM (llama-3.3-70b-versatile) handling natural language FAQ responses.

PILLAR 02
Custom CRM Platform

A bespoke patient record management system pre-loaded with 1,888 existing patient records. Features advanced search, multi-column filtering, Lizto-compatible CSV export, 50-record pagination, and direct WhatsApp contact per patient. Standalone HTML — no server, no database, no ongoing maintenance overhead.

PILLAR 03
Corporate Digital Presence

A full-featured corporate website built to the clinic's visual identity (cream/gold/brown palette, Playfair Display + DM Sans). Includes service catalog, video production banner, embedded WhatsApp CTA, appointment form, and Fibot integration entry points. Deployed to Cloudflare Pages globally.

PILLAR 04
Lizto CRM Integration & Analytics

A dual-path data pipeline connecting Lizto's appointment management system to the custom CRM via Make automation. Path A intercepts booking confirmation emails. Path B provides a CSV export/import bridge. A centralized analytics dashboard (in development) will unify Meta Ads, Lizto, and bot metrics.


04 · Fibot AI Chatbot Suite

Architecture, Conversation Design & Deployment

Each Fibot instance operates as a hybrid: a deterministic finite-state automaton governs core flow steps, while the Groq LLM handles open-ended FAQ responses. This eliminates hallucination risk in critical patient-facing steps while preserving conversational fluidity.

Standard Bot Flow — 6 Steps
  1. Greeting & brand introduction (Fibot identity, Dr. Esteban Pinto)
  2. Antecedents question: prior procedures? [Si / No — hardcoded buttons]
  3. If Si: time reference capture (regex-detected free text)
  4. Photo request: front + profile (2 images required, reference image shown)
  5. Name collection (blocklist rejects interrogative words)
  6. WhatsApp advisor card — lead delivered, pipeline complete
FAQ Interception Pattern
Fires at Any Step

When a patient asks a question at any flow step, the FAQ interceptor fires. Groq LLM resolves the query. repeatCurrentStep() returns the patient to their exact position. The pipeline is never interrupted.

Deployed Service Bots
Alectomía
Perfilamiento Labial
Rinomodelación
Botox Full Face
Bichectomía
Correc. Ojeras
Correc. Surcos
Hydroxyfill
Jalupro 5.5

05 · CRM & Data Strategy

Custom CRM, Patient Data Management & Lizto Integration

Custom CRM System
1,888 Patient Records · Standalone HTML

Bespoke patient record management built as a standalone HTML application — no server, no database, no maintenance overhead.

Full-text searchMulti-column sort50-rec paginationCSV Lizto exportWhatsApp per recordNo backend
Lizto Integration — Dual Path
Path A · Email Interception (Primary)

When Lizto dispatches a booking confirmation email, Make's Email Watch module intercepts it, parses structured data fields, and routes the record to the Custom CRM via HTTP connector — near-real-time, zero manual steps.

Path B · CSV Export Sync (Secondary)

Lizto's CSV export is used for historical migration and periodic bulk synchronization. The Custom CRM includes a Lizto-compatible import layer with field normalization and deduplication.

Status
70% Complete

Both paths validated. Real-time bridge in final configuration.


06 · Technical Infrastructure

The Zero-Infrastructure-Overhead Technology Stack

The core architectural principle: elimination of dedicated server infrastructure. Every component operates on stateless, serverless, or edge-deployed platforms — zero monthly hosting fees, zero DevOps overhead, global CDN distribution from day one.

01
Presentation Layer
HTML5 · CSS3 · Vanilla JS

Single-file bot architecture. No framework, no build pipeline. Deployable as flat ZIP. Maximum portability, zero dependency management.

02
AI Inference Engine
Groq · llama-3.3-70b

High-throughput LLM inference for FAQ resolution. Hardcoded state-machine handles deterministic flow steps without API dependency. Sub-300ms latency.

03
Edge Deployment
Cloudflare Pages · Global CDN

300+ global edge locations. Sub-100ms TTFB. Zero infrastructure overhead. 99.99% uptime SLA. No SSL management, no bandwidth billing.

04
Automation Engine
Make (ex-Integromat)

Multi-step workflow orchestration for Lizto integration. Email parsing, HTTP connectors, error handling + retry logic. Visual modification without code changes.

05
CRM Data Layer
Custom CRM + Lizto

1,888 patient records. Lizto-compatible CSV. WhatsApp direct-dial. Dual-path sync. No backend dependency on either system.

06
Paid Media Layer
Meta Business Suite

One campaign per service vertical, one Fibot per campaign. Direct ad-to-bot routing. No intermediary landing page friction. Full attribution pipeline.


07 · Meta Ads Integration

How Paid Media Connects to the Automation Ecosystem

The one-campaign-one-bot principle: each of the nine active service verticals maintains a dedicated Meta Ads campaign delivering traffic directly to its corresponding Fibot instance — eliminating service confusion and maximizing lead qualification specificity from first interaction.

The Lead Journey: Ad Impression to Advisor Derivation

  1. Patient sees a service-specific Meta Ad — creative qualifies intent at the scroll level.
  2. Clicks through to the Fibot landing page on Cloudflare — nearest edge node, sub-100ms load.
  3. Fibot delivers greeting and service intro, establishing trust via Dr. Esteban Pinto branding.
  4. Antecedents question: prior procedures? (hardcoded Si/No buttons). If Si: time reference collected.
  5. Two reference photographs requested (frontal + profile). Upload zone + reference image displayed until both received.
  6. Patient name collected and validated (blocklist rejects interrogative patterns).
  7. WhatsApp advisor card delivered — specialist receives pre-screened lead with all data captured.

08 · Results & Business Impact

Measurable Outcomes Delivered

9
AI BOTS IN PRODUCTION
One per active Meta campaign. Deployed to Cloudflare global edge. Operating continuously since launch.
24/7
UNINTERRUPTED LEAD CAPTURE
All hours, all days. Zero human intervention during non-office hours. Leads qualified and routed immediately.
100%
PRE-SCREENING AUTOMATION
Every lead arrives with clinical history, photos, and service context already captured. Advisor preparation time: zero.
70%
CRM INTEGRATION COMPLETE
Lizto data pipeline operational via Make. Both paths validated. Final bridge configuration in progress.
Operational Efficiency

Advisor intake time eliminated. Fibot handles 100% of pre-qualification. Every WhatsApp derivation arrives with a complete, actionable lead package.

Infrastructure Overhead

Total hosting cost of 9 bots + CRM + website: zero per month. Zero DevOps overhead. Global CDN performance with fully optimized operational margins.

Scalability

2-hour deployment per new service bot. Bot madre architecture scales linearly to the full 73-service roadmap without structural refactoring.


09 · About the Firm

Built by Tribo

Tribo is a boutique automation and AI integration studio specializing in end-to-end digital infrastructure for service businesses. We deliver conversational AI systems, CRM pipelines, and marketing automation ecosystems with production-grade quality and measurable business outcomes.

AI Chatbot Development

WhatsApp bots · Web chatbots · Lead capture automations · LLM integration · State-machine design · Multi-service suites

CRM & Automation

Custom CRM systems · Make workflows · Data pipeline architecture · Multi-platform synchronization · CSV/API bridging

Meta Ads Integration

End-to-end paid media to bot to CRM pipelines · Campaign-specific bot architecture · Lead quality optimization