5+ years
AI & Data Science experience
Available for senior AI roles
Building production AI systems that create measurable business impact.
AI Engineer with a master's degree and 5+ years of hands-on experience designing, developing, and deploying machine learning and generative AI solutions. Proven track record across insurance, retail, and energy with end-to-end ownership from data pipelines to cloud deployment.
Global Teams
Proven delivery experience for international companies and cross-cultural stakeholders
Cross-Industry Impact
Proven ability to design and deploy AI solutions adaptable to any industry domain
Representative work
Project Portfolio Highlights
AI Model for Estimating the Labour Impact of FNCERs (UPME, 2025)
Research publication focused on building an AI-based model to estimate and analyze the labor-market impact of non-conventional renewable energy sources in Colombia.
- Published with UPME (Unidad de Planeación Minero Energética), January 2025.
- Applied AI methodologies to evaluate workforce impact scenarios for energy transition planning.
Optimization of Residual Heat from Internal Combustion Engines Through ORC (2021)
Technical research on thermodynamic and economic optimization of Organic Rankine Cycle (ORC) systems to improve energy recovery from internal combustion engines.
- Published in the Colombian research network, June 2021.
- Focused on optimization and performance modeling for energy efficiency applications.
PolicyFit Advisor
Live DemoProduction-ready full-stack platform that uses advanced AI and LLM workflows to analyze commercial insurance policies, identify critical coverage gaps, and generate tailored recommendations aligned with business profile context.
- Core AI features: policy extraction, gap detection, recommendation scoring, and analysis chat per request.
- Backend: FastAPI, async architecture, OpenRouter/OpenAI-ready LLM client, PDF/DOCX processing, and request tracing.
- Frontend: React + Vite with 7 feature pages, Zustand state management, and optimized production performance.
Chemical Report App
Live DemoFull-stack AI application to generate, manage, and analyze chemical reports with a production-ready architecture and role-based access.
- Monorepo setup: FastAPI backend (`apps/api` on Render), user web app (`apps/web` on Vercel), and admin panel (`apps/admin` on Vercel).
- Core capabilities: JWT auth with roles, LLM-powered report generation, report persistence in Neon PostgreSQL, professional PDF generation, and report-level chat.
- User experience: report history, markdown-enhanced visualization, panel-based layout, and full-space responsive interface.
- Admin operations: users list, all reports list, per-user reports view, protected login, and centralized API integration.
Energy Regulatory Copilot
Live DemoAI-powered assistant specialized in Colombian energy regulatory matters. This full-stack app combines RAG with LLMs to deliver accurate, context-aware answers and supports chat history, document indexing, semantic retrieval, and streaming responses.
- Backend: FastAPI, Chroma vector DB, OpenRouter, PyPDF, REST + streaming.
- Frontend: React 18 + Vite, Markdown rendering, React Hooks state management.
- Multi-mode chat: general QA and document-grounded RAG queries.
Additional projects in ML, NLP, analytics, and deployment are available in my full GitHub portfolio: Data-Related-Projects.
Career path
Professional Experience
Senior AI Developer · Linqura
Aug 2024 - Present- Designed and deployed a multi-agent classification system for commercial insurance using web scraping, vector databases, FastAPI, and LangChain orchestration.
- Led development of RAG-based document intelligence for policy analysis and claims automation workflows.
- Managed LLM serving on AWS using vLLM with OpenAI-compatible endpoints for scalable production inference.
- Led SLM fine-tuning with Unsloth for NAICS keyword generation, improving classification quality by approximately 25-30%.
Data Scientist · The Home Depot Canada
Apr 2023 - Aug 2024- Fine-tuned and deployed BERT-based classifiers for reviews and service feedback, reducing manual labeling effort.
- Built end-to-end NLP pipelines in Python and BigQuery for support chats and customer reviews.
- Implemented sentiment analysis, topic modeling, and intent classification for decision-ready dashboards.
- Developed ETL pipelines and custom scripts for ingestion, cleaning, transformation, and analytics automation.
Data Scientist · DevSavant
Jul 2022 - Mar 2023- Built interactive BI dashboards in Power BI and Tableau for real-time business visibility.
- Created Python automation workflows for extraction, processing, and reporting at scale.
- Integrated cloud workflows and AI models with third-party APIs for clients in logistics, healthcare compliance, and AI communications.
AI Developer · Neu Energy
Oct 2021 - Jun 2022- Developed ML models for energy disaggregation and forecasting to optimize clean-energy trading strategies.
- Built backend APIs in FastAPI for high-volume energy data processing.
- Worked with PostgreSQL, AWS Athena/S3, and Docker for model and service deployment.
Junior Data Scientist · Allianz Energy
Feb 2021 - Sep 2021- Built Python automation tools to reduce manual reporting and improve data quality.
- Developed simulation models for solar resource performance and investment planning.
Technology stack
Core Skills
Machine Learning & GenAI
TensorFlow, PyTorch, Scikit-learn, XGBoost, LightGBM, Hugging Face, LangChain, LlamaIndex, BERT, RAG, GPT fine-tuning
MLOps & Deployment
vLLM, Ollama, FastAPI, Flask, gRPC, Pinecone, Weaviate, FAISS, ChromaDB, AWS SageMaker, GCP Vertex AI, Kubernetes, Docker
Data & Engineering
Python, SQL, Bash, BigQuery, PostgreSQL, ETL pipelines, API integrations, GitHub Actions, Tableau, Power BI
Academic profile
Education, Certifications & Publications
Education
- MSc in Artificial Intelligence · Universidad Internacional de La Rioja (Jun 2022 - Jan 2024)
- BSc in Energy Engineering · Universidad Autónoma de Bucaramanga (Jun 2017 - Jul 2021)
Professional Certifications
- IBM Data Science Professional (Aug 2021 - Mar 2022)
- IBM AI Developer Professional (Nov 2023 - May 2024)
- IBM RAG and Agentic AI Professional (Apr 2025 - Aug 2025)
Selected Publications
- An artificial intelligence model for estimating and analysing the labour impact of FNCERs (UPME, Jan 2025)
- Energy cost modeling in Colombia: A comparative approach to AI-based solutions (UNIR, Sep 2023)
- Optimization of residual heat from internal combustion engines through ORC (Colombian research network, Jun 2021)