01
Hello
Data & Automation Engineer
Santa Cruz de la Sierra, Bolivia
Tools: Python, n8n, Apache Airflow, dbt, PostgreSQL, Snowflake, BigQuery, Docker, FastAPI, LangChain, WhatsApp API, Power BI
Built systems for
- CarSans
- Group Quimera
- Gerona SVF
- US e-commerce
- US home automation
Four things I do properly
50%
Faster data loads. Parallelised ETL with automated validation at Gerona SVF.
40%
Less time per lead. WhatsApp → CRM → Postgres, synced in real time.
95%
Report reliability. Cleansing, normalisation and validation rules.
3+
Years shipping. Bolivia and the US, in Spanish and English.
Anatomy of one real automation
The lead engine I built at Group Quimera. A message lands on WhatsApp and no human touches it until a salesperson is pinged. Follow one lead through it.
Step 1 of 5/WhatsApp Business API
WhatsApp: message in
A prospect writes to the business number. The WhatsApp Business API hands the message to the workflow the moment it arrives.
Sample data
CRWhatsApp
+591 7•• ••• 412
Hi! I saw your ad and I'd like more information and prices, please.
10:42
Four steps, zero guesswork
A short, measurable process. You know what happens at each stage and what you get at the end.
01Diagnose
I map your operation and find where hours and money actually leak.
Week 102Design
The full flow: what it does, what it touches, how the result gets measured.
Week 1–203Build
Built, integrated and tested against real data before it reaches production.
Weeks 2–404Tune
Monitored and scaled. You get reports with hours saved, not vibes.
Ongoing
Things I have built
Real systems, running in production, with results you can measure.
Hi, I'd like info and prices
10:42
WhatsApp
n8n
CRM
Postgres
Sales
Lead #1284 → Ana
AutomationGroup Quimera · 2025
- less processing time per lead
- 40%less processing time per lead
WhatsApp lead automation engine
- The problem
- Every WhatsApp lead was copied into the CRM by hand, and some were lost.
- What I built
- An n8n workflow that captures, validates, syncs and routes each lead in real time.
- n8n
- WhatsApp Business API
- PostgreSQL
- REST APIs
- +1
PIPELINES
OK
LOAD TIME
-50%
RELIABILITY
95%
DataGerona SVF · 2024
- faster load times
- 50%faster load times
- report reliability
- 95%report reliability
Corporate data warehouse and ETL pipelines
- The problem
- Sales, inventory and CRM lived apart, and reports were assembled by hand.
- What I built
- Airflow and dbt pipelines into Snowflake and BigQuery, with Power BI on top.
- Python
- Apache Airflow
- dbt
- Snowflake
- +2
Casera
Fiado
Bs 165
Doña RosaBs 120
Don JulioBs 45
MartaBs 0
Own product2026
- works with no internet and no account
- Offlineworks with no internet and no account
Casera: digital ledger for market vendors
- The problem
- Market vendors keep sales and customer credit in a paper notebook.
- What I built
- An offline-first app for cheap Android phones that tracks who owes what.
- React
- Vite
- PWA
- TypeScript
- +1
Three processes that stopped being manual
Real work from companies I worked with. Flip the switch to see how each one ran by hand, and how it runs now.
Showing how each process ran by hand.
- –Done by hand
Every WhatsApp lead was copied into the CRM by hand and classified by whoever was free.
Group QuimeraLead intake
- –Done by hand
Reports were stitched together by hand from three separate systems, and nobody fully trusted the numbers.
Gerona SVFReporting
- –Done by hand
Data loads ran one after another for hours, with no way to know when something had failed.
Gerona SVFData pipelines
Got a process that should run itself?
Tell me what is eating your team's time. If I can help, I will tell you how; if I cannot, I will tell you that too.
Send an email