Henry Rincon

I lead the ISO/IEC 42001 AI strategy and the analytics of a healthcare institution.
• AI management system: 20+ instruments, including the systems inventory, risk matrix and impact assessment.
• 25+ Power BI analytics products, with their data quality and privacy rules, used by some 20 leaders across about 15 processes.
• AI agent prototypes with LangChain, LangGraph and n8n.

I design and build my own generative AI systems, end to end.
• A public RAG assistant on my résumé app that cites its sources, evaluated on 77 questions of my own and 148 external ones.
• ARKHÉ, an ecosystem of specialized agents: 52% fewer tokens than a generalist agent over 120 scenarios.
• 13 agents on Claude Code, including an ISO/IEC 42001 expert and a Power BI builder over MCP.

I designed the data strategy and architecture of an automation-agents startup, working directly with its founders.
• A Microsoft Fabric ecosystem from scratch: 12 clients, 120 tables, 20 GB, about 1,000 events a day.
• Real-time monitoring of up to 23 agents in production, out of 27, with their n8n flows integrated into Fabric.
• Power BI layer with Direct Lake and row-level security.

I co-led the area's data governance and led its business intelligence and predictive analytics.
• Five-person BI team: Power BI dashboards for 50+ business users and a 25% improvement in decision-making.
• Churn, delinquency and risk models in Python and scikit-learn, above 90% accuracy, delivered in Power BI.
• Power Query and ETL optimization: about 35% shorter analysis times.

I led advanced analytics and data transformation for TransMilenio's post-operational analysis.
• Power BI rollout adopted by 25+ key users, with 35% more efficient analytics processes.
• ETL integrating five sources, from fare collection to fleet GPS, with explicit quality rules.
• Demand forecasting by route and time slot in Python and scikit-learn, still in use after I left.

I led the testing and validation of Oracle WMS Cloud in the distribution center's transition.
• A mixed team of 20 people for about six months: 25% fewer errors and 15% more operational efficiency.
• The BI to control the rollout, adopted by 15+ senior decision-makers.
• VBA integrations with 50% fewer data errors, and SQL data quality control between systems.

I built the operational analytics for the supervision of TransMilenio's routes.
• Performance reports on about 150 routes of 10 concessionaires, from five sources, used by 20+ internal users.
• VBA, Power Query and SQL automation of the weekly data: 30% less processing time.
• A historical repository in SQLite, my first end-to-end data pipeline.

I led analytics and process transformation for an educational assessment operation covering 100+ schools a year.
• Scheduling, logistics and digitization KPIs, presented weekly to the board of directors.
• Close to 40 people coordinated across three fronts.
• A VBA optimizer to assign test administrators under explicit constraints, and the move to Scrum, Kanban and BPMN.

I led the operational and technological transformation of a heavy-machinery operation with about 120 units.
• Odoo ERP implementation, and a team of 12 technicians and operators with 95% customer satisfaction.
• The SQLite analytics database the operation did not have.
• Process modeling in BPMN and simulation in FlexSim before changing the operation.

I optimized processes with statistical analysis, Lean Six Sigma and operations research.
• KPI dashboards to track operational performance and prioritize improvements on evidence.
• Improvements in inventory, shift assignment, process costing and the layout of equipment and spaces.
• Processes standardized in Bizagi and simulated before rollout; VBA and Power Pivot tools.

Grade: 4.5

Microsoft certification for designing, building and administering end-to-end analytics solutions on Microsoft Fabric: data preparation and transformation, lakehouses and warehouses, semantic models, and data protection and governance, with SQL and DAX.

Applied data science with R: statistical analysis, predictive modeling, hypothesis testing, SQL integrated with R, visualization with ggplot2 and interactive apps with R Shiny, in RStudio, with a final project on real data.

Grade: 4.7

Professional certificate in data science: data preparation and exploration, analytical programming in Python, SQL queries on relational databases, visualization and model building with pandas, NumPy and scikit-learn.

Grade: 4.7

A course of the IBM Data Science Professional Certificate: Python for data science, from preparing and exploring data to analysis, with pandas and NumPy.

A course of the IBM Data Science Professional Certificate: SQL on relational databases to locate, relate, filter and validate the data before any analysis, with queries and subqueries.

Grade: 4.7

Professional certificate in data science: data preparation and exploration, analytical programming in Python, SQL queries on relational databases, visualization and model building with pandas, NumPy and scikit-learn.

Grade: 4.5

Course in linear optimization at Universidad de los Andes. I apply it in integer programming with Pyomo and HiGHS, constraint programming with OR-Tools, and assignment under explicit constraints.

Grade: 4.5

Course in probability and statistics at Universidad de los Andes. I apply it in my predictive models: comparison against a baseline, confusion matrix and temporal validation with RMSE.

Grade: 4.0

Activities and societies: Industrial engineering with an emphasis in data analytics intelligence: process optimization, data analysis and decision support, with modeling and forecasting of industrial phenomena using SQL and scikit-learn, Python and Power BI.

Bachelor's degree in Industrial Engineering with a focus on Analytical Data Intelligence.

Intensive English course and IELTS certification.