Oscar Tibaduiza

Technical Projects and Achievements:

MCP (Model Context Protocol) Agents Development: Creation of MCP servers in Python (FastMCP) to connect LLMs (Claude/OpenAI) to local environments and databases (System Monitor, SQLite Explorer).

RAG Architecture & Streamlit: Design of an "AI Data Engineer Assistant." An interactive Streamlit application capable of analyzing GitHub repositories, explaining Airflow pipelines, and optimizing SQL queries in real-time.

End-to-End Machine Learning: Deployment of a suite of 5 ML models (Churn, NLP, Time Series with Prophet) integrated into a unified web interface for client demonstrations.

Technical Training: Creation of advanced courses on SQL automation and the use of NotebookLM for automated technical documentation.

Execution of 7 major projects covering the complete Data Science lifecycle, in various business contexts (Finance, E-commerce, Big Data).

1. Big Data & Cloud (AWS/Spark):

Deployment of a Big Data architecture on AWS cloud (EMR).

Execution of a distributed image processing pipeline with PySpark.

2. NLP & Computer Vision (E-commerce):

Automatic classification of consumer goods for a marketplace.

Textual analysis (NLP): Tokenization, TF-IDF, Word2Vec, BERT.

Image processing: Feature extraction (SIFT) and Transfer Learning (CNN).

3. Modeling & Scoring (Finance):

Development of a scoring model to predict loan repayment probability.

Creation of an interactive dashboard to explain model decisions to account managers (Streamlit/Dash).

4. Customer Segmentation (Marketing):

Exploratory data analysis and cleaning (EDA).

Unsupervised clustering (K-Means, DBSCAN) to identify distinct customer profiles.

Optimization of model maintenance frequency.

Technical environment: Python, AWS (S3, EC2, EMR), Spark, TensorFlow, Git, Jupyter.

Management of critical incidents and technical support for high-performance storage infrastructures (SAN/NAS) for large accounts.

Key Responsibilities:

Support & Troubleshooting (L2/L3): Remote resolution of configuration, performance, and hardware failure incidents on Hitachi storage arrays (VSP Gx00, HUS, AMS).

Infrastructure & Virtualization: Management of virtualized environments (VMware, Hyper-V) and SAN switches (Brocade, Cisco).

Incident Management (ITIL): Analysis, qualification, and escalation of tickets (Hi-Track tools). Coordination with global teams (Global Support Center) for critical cases impacting production.

Maintenance: Supervision of "Health Checks," firmware updates, and security patches.

Environment: Storage (SAN/NAS), Hitachi Vantara, Cisco, Brocade, VMware, Windows Server, Linux.