AI IMPLEMENTATION CONSULTANT
I ran the operation. Now I build the systems that run it.
7+ years operating two healthtech companies in Mexico, up to general management. Since 2022 I implement AI and automation inside service businesses — from diagnosis to a team that runs it on its own.
- Anthropic-certified
- Airtable-certified
- Remote from Mexico (GMT−7)
Open to full-time remote roles — AI implementation, AI consulting, GTM.

Results from my own operation
Source: Dentrilab · 2023–2026
Three steps, in this order.
I start where the operation loses time and money, build on the tools it already uses, and stay until the team runs the system without me.
Leaks mapped on the real process 01Diagnose
Where time and money leak, step by step, plus a data audit: what the data can support today and what it cannot yet.
A new layer over the existing tools 02Build on what you already have
The system goes on top of the tools the operation already runs. No system replacement.
Before and after, against a baseline 03Adopt and measure
A supervised pilot with a human in the loop, before/after against a baseline, and training for the team.
One operation of my own, two engagements with others.
Each case follows the same template: problem, before, architecture, after and what is replicable. Where a metric is not verified yet, the case says so.
Back-office in three tech generations
The back-office of my own company moved through three tech generations in six years, while the company kept operating under COFEPRIS regulation.
Read the case- 012020–22
- Power PlatformGeneration one
- 022022–25
- AirtableGeneration two
- Make · ZapierAutomations
- 032025–26
- Supabase · n8nGeneration three
- Agent orchestrationGeneration three
- Vertical SaaS ERPNext.js · Supabase · Vercel
- X-ray reading pilotInternal · no diagnostic use
- 04Result
- 25 h/week removedTwo automations
- +21% gross marginInventory and delivery
National medical-equipment distributor
A process diagnosis and a process playbook for a national distributor of medical equipment. Both delivered.
Read the caseConstruction company, inside a consultancy’s team
A one-week paid diagnostic for a construction company, working inside a consultancy’s team. It ended in an ERP scope and a statement of work.
Read the caseSix sandboxes you can operate.
Each one runs in the browser on synthetic data. Three mirror systems that ran in my own operation; three are concepts. The tag says which.
Procedure Unit Economics
A lab can grow revenue and lose money. This shows which products actually pay for the technician hours they consume.
Open the demoOps Dashboard with auto-reports
Owners open a dashboard, see twelve widgets and close it. This one writes the summary in sentences instead.
Open the demoVertical mini-ERP
Most service businesses need four or five linked tables, not forty generic SaaS objects. This is that data model, navigable.
Open the demoSandbox demos · synthetic data · no API calls at runtime
All demosThree ways in.
White-label
For consultancies: implementation with a consultant’s judgment.
For software development firms: the operator in the room — discovery, process and adoption next to your developers.
Operator first, builder since 2022.
Two healthtech companies co-founded and operated in Mexico. Independent AI implementation consulting since 2022.
- 012022 – presentIndependent consulting
Principal AI Implementation Consultant · Cauce
Independent AI implementation consulting since 2022, operating as Cauce since 2026. Ran in parallel with Dentrilab until April 2026, full-time since. I have worked with consultancies and construction companies.
- Development subcontracting with Sonora Cloud.
- 022020 – April 2026Dentrilab
Co-founder · General Director
Program Manager 2020–2023, General Director 2023–April 2026.
- $4–5M MXN annual revenue, 12–15% yearly growth, teams of 9–15, COFEPRIS-regulated.
- 25 h/week of manual admin removed with two automations, in my own operation.
- +21% gross margin via inventory optimization and faster delivery.
- Three tech generations: Power Platform (2020–22) → Airtable/Make/Zapier (2022–25) → Supabase/n8n/agent orchestration (2025–26).
- Vertical SaaS ERP (Next.js, Supabase, Vercel, Claude Code).
- Internal AI-assisted preliminary X-ray reading pilot, with legal disclaimer, no diagnostic use.
Source: Dentrilab · 2020–2026
- 032017 – 2019Montelo 3D
Co-founder · Product & Clinical Adoption Lead
Started during university.
- $1.2M MXN seed.
- Won, with the team, the first tender for 3D surgical-planning models in Mexico (IMSS La Raza).
- Led adoption with sub-specialist surgeons.
Source: Montelo 3D · 2017–2019
- 042016GRUPO JOIN
Strategic Planning Consultant
Speaking
- Speaker, Biomedical Engineering Weeks (2020–2025).
- Speaker, “La nueva era del trabajo: Personas, Tecnología e IA”, HeyWork Los Mochis (2026).
What I run in production.
AI & build
- Claude Code
- Claude API (MCP, SDK)
- Agent orchestration
- Next.js
- Supabase
- Vercel
- PostgreSQL
- Python
Automation & data
- Airtable
- n8n
- Make
- Zapier
- REST/webhooks
- Power Automate
- Power BI/Tableau
Operations
- Process diagnosis
- Data audit
- Supervised pilots, human in the loop
- Before/after against a baseline
- Team training and adoption
- COFEPRIS-regulated operation
Certifications
Anthropic
- AI Fluency
- Subagents
- Claude Code 101
- Claude Platform 101
- Deploying Claude Enterprise with Confidence
Airtable
- Builder
- Admin
- AI-First Foundations
- Agentic Systems Design
Honors
- 1st place national, IMSS Health Innovation Olympics, 3rd edition (2018).
- Nominated, Innovators Under 35 LATAM, MIT Technology Review (2019).
Education
Biomedical Engineering · Universidad La Salle Noroeste · 2014–2019
Thirty minutes, no deck.
Pick a time, or write to me. I answer from Los Mochis, Mexico (GMT−7).
- linkedin.com/in/geraerodriguez
- CV
- New version in preparation
- Loom
- Watch the intro Loom · 90 s
Book 30 minutes
Open in Cal.comThe calendar loads from Cal.com.