Process diagnosis & automation
I map how work actually flows, quantify where time and money leak, and rebuild the process with automated pipelines instead of spreadsheets and copy-paste.
- Python
- n8n
- Make
- APIs & Webhooks
- RPA
- FlexSim
Industrial engineer and founder. I diagnose how organizations actually work, quantify the inefficiency, and rebuild the process around data and automation — cutting up to 90% of manual processing time for clients in manufacturing, energy and construction.
Selected and recognized in 2026 by
Reduction in manual processing time delivered to clients
Projected campaign ROI lift from the MercadoLibre segmentation study
Row database built and analyzed with SPSS multivariate methods
Finalist among 500+ participants — 30X × Colsubsidio Hackathon
Every engagement starts the same way: understand the real process before automating it. Then build the smallest system that removes the bottleneck for good.
I map how work actually flows, quantify where time and money leak, and rebuild the process with automated pipelines instead of spreadsheets and copy-paste.
From raw, unstructured operational data to dashboards leadership actually uses: structured datasets, statistical analysis and reporting that answers the decision, not just the question.
Agentic workflows and RAG pipelines built on the Anthropic and OpenAI APIs — document extraction, conversational advisors and assistants wired into the systems a team already uses.
Infrastructure on Stellar that makes donor funding traceable and verifiable for NGOs. Foundations capture expenses from a photo through a WhatsApp bot, funds are disbursed and anchored automatically, and donors verify how every peso was used — without ever seeing a wallet or a hash. Selected participant at the Stellar Builder Summit and Stellar House, São Paulo.
End-to-end automation and data work for operations-driven companies across manufacturing, energy and construction. I lead each engagement from problem scoping to delivery: diagnose the process first, then build reporting pipelines, dashboards and internal web tools that replace the manual routine entirely.
Built a 5,000-row primary research database and ran multivariate analysis in SPSS — cluster analysis, ANOVA and regression — to profile consumer behavior. Isolated three high-value segments and translated them into a go-to-market strategy for the marketing team.
Led a benchmarking study across oil & gas, construction and manufacturing: competitor strategies and product portfolios mapped through primary research and industry interviews, then turned into structured datasets, Power BI dashboards and executive reports used for leadership-level decisions.
An LLM advisor that automates an end-to-end insurance sales process: it qualifies the customer, explains coverage in plain language and carries the conversation to closing. Finished 5th out of more than 500 participants.
Built and pitched under 48-hour constraints, against teams from across the Americas.
International hackathon held in Argentina's wine region, bringing together builders from across LATAM.
International competition across the Americas hosted by the Stellar blockchain ecosystem.
Won at the Stellar Builder Summit, engaging directly with Stellar leadership and the LATAM ecosystem.
Conversational insurance advisor automating an end-to-end sales process, built in under 48 hours.
Research on anchoring building-information models to a verifiable ledger so that project data — quantities, changes, approvals — can be audited without trusting a single party. Published by Springer and presented at the Colombian Computing Congress.
Ongoing work at the intersection of engineering education, computing and verifiable data — the same thread that runs through everything I build: if a number matters, it should be possible to check it.
Analytical rigor on one side, shipping the solution on the other.
Pontificia Universidad Javeriana
Bogotá · expected 2027
Trailhead
2026
Codecademy
2025
MinTIC Colombia · + BI Data Analyst, Codecademy (2024)
2024
Tell me what part of the process still runs on spreadsheets and manual copy-paste. I'll tell you what it costs you and what it would take to remove it.