Projects

Selected implementations showcasing AI, automation, and data science solutions. All projects maintain confidentiality with respect to client information.

Hybrid PII Detection System

Insurance

LATAM + Malaysia 100% compliant Real-time alerts

High-precision PII detection using hybrid regex + transformer architecture for multi-country regulatory compliance.

Hybrid NLP Transformers Compliance Python

View Details

OCR-Based Medical Claim Fraud Detection (Chile)

Insurance

4,562 hrs/year saved $840K/year saved 1K prescriptions/day

Daily batch processing system for Chilean medical claims using pytesseract OCR, Sentence Transformers, and FAISS to identify duplicate prescriptions and detect fraud patterns across scanned medical documents.

pytesseract OCR Sentence Transformers FAISS Medical Claims

View Details

RPA Pension Payment Automation System

Insurance

50K hours saved/year 500K+ PDFs/month Zero IP blocking

Automated pension payment processing with anti-blocking web scraping, PDF extraction, and SharePoint integration.

RPA Python Selenium SharePoint

View Details

Agentic AI Framework

Insurance

70% less documentation 40% faster migration 30% better code quality

Standardized programming practices across LATAM using GitHub Copilot and autonomous AI agents for code review, migration, and optimization.

GitHub Copilot Agentic AI Code Migration Python

View Details

Analytics Dashboard & Backoffice Process Optimization

Retail

40% faster validation 25% fewer fines Data-driven ops

Data-driven workflow redesign and compliance dashboards for retail operations management.

SQL Dashboards Data Analysis Process Optimization

View Details

Customer Service Excellence & KPI Management System

Telecom

86% satisfaction 24% fewer complaints 120 executives

KPI-driven team management and quality assurance for customer service excellence in contact centers.

KPI Monitoring Team Management Service Analytics Quality Assurance

View Details

Predictive Analytics & Competitive Intelligence Platform

Retail

44% → 49% productivity Real-time competitor tracking ML-driven insights

ML models for customer retention, churn prediction, and real-time competitor pricing intelligence.

Scikit-learn Web Scraping Churn Prediction Python

View Details

Technical Approach

Agentic AI Architecture

I design autonomous systems using agent frameworks (LangChain, AutoGen) that can reason, plan, and execute complex tasks. These systems combine LLMs with tools and memory to create intelligent workflows.

Hybrid ML Systems

My approach combines the strengths of different techniques: transformers for semantic understanding, rule-based systems for determinism, and traditional ML for specific tasks. This ensures reliability and performance.

Data Quality First

I prioritize data quality and pipeline robustness. Every project includes comprehensive data validation, quality monitoring, and automated alerts to ensure system reliability.

Business Alignment

Technical solutions are always aligned with business objectives. I focus on measurable KPIs and ensure every feature drives actual business value.