Machine Learning • Computer Vision • NLP • Healthcare Analytics • BI • Cloud

I'm a Data Scientist and Biomedical Engineer with experience applying AI models, automation and advanced analytics in fields like healthcare, scientific research and business intelligence. I blend technical skills with analytical insight to build end-to-end data solutions, from pipelines and predictive models to strategic dashboards and intelligent agents.

• Predictive models, classification and time series
• Computer Vision (BEiT, UV simulation, embeddings, filters)
• NLP, LLMs and autonomous agents
• Experiment optimization and pipeline tuning

• Advanced dashboards in Power BI, Tableau, Looker Studio
• Semantic models, DAX, SQL
• Strategic visualization and data storytelling
• Report automation

• Clinical analytics, cohorts and biomarkers
• Medical image processing
• Technical-scientific support for research projects

• ETL with Python, Apps Script and Google Cloud
• APIs and workflow automation
• Streamlit, CI/CD and containers

Python • SQL • R • JavaScript

Scikit-Learn • PyTorch • TensorFlow • BEiT • HuggingFace
LangChain • LlamaIndex • LLM Agents

Airflow • Google Cloud Run • REST APIs
Pandas • NumPy

Power BI • Tableau • Looker Studio
Plotly • Seaborn • Matplotlib

Docker • Git • Streamlit • GCP • AWS (basic)

📋 Recognizing birds by adding a synthetic UV channel and extracting embeddings with BEiT.

🔗 Repo:
https://github.com/Vagarh/Sistema-de-Visi-n-por-Computadora-para-Reconocimiento-de-Aves-y-Simulaci-n-Tetrocrom-tica

Results:

• Accurate tetrachromatic vision simulation
• Embeddings → UMAP → species clustering
• Complete pipeline: cleaning → normalization → feature extraction

📋 Automating processes via autonomous agents.

🔗 Repo:
https://github.com/Vagarh/Agentes-de-IA-con-Modelos-de-LLM

Includes:

• Conversational agents
• Text classification
• Appointment management
• API integration

📋 Agent that generates personalized plans for running, cycling and triathlon.

🔗 Repo:
https://github.com/Vagarh/Endurance-Lab-AI-Desaf-o-de-Evaluaci-n-Autom-tica-Agente-Entrenamiento-Deportivo

Features:

• Recommendations for load and zones
• Dynamic adjustments based on fatigue
• Progress analysis

📋 Advanced vision projects and ML.

🔗 Repo:
https://github.com/Vagarh/Procesamiento_Imagenes_curso_eafit_2025

Includes:

• HOG
• SVM
• Clustering
• Colorimetry
• Image normalization

🎨 Interactive portfolio:
https://studio--datadive-portfolio-d7zco.us-central1.hosted.app/#home

Includes:

• Analytical dashboards in Power BI, Tableau and Looker
• Automated ETL with Apps Script + Python
• Risk models, PQRS, clinical cohorts, etc.