Ismael Bovingdon Castillejo

Data Engineering | Cloud | Business Intelligence

I build data pipelines that hold up in production

4+ years of production SQL on high-volume financial systems at BBVA, and an internship with Azure Data Factory, SQL Server and Power BI at Babel. MSc in Data Science (UCM, 2026), awarded 1st Prize for a thesis built on a 24/7 production data capture.

Get in touch Download CV Open my thesis app

Available to start now · Madrid, on-site or hybrid

About me

Professional focused on data, cloud and BI

I started at BBVA Technology: 4+ years of SQL on DB2 in high-volume financial systems, owning the quality of 3 batch chains with accounting impact and cutting the runtime of slow processes by up to 50%.

I planned my move into industry by taking a Master’s in Occupational Health & Safety during my last year at the bank. Within a year and a half I became Health & Safety Manager, coordinating 20+ workers across 4 parallel sites for Canal de Isabel II, and I backed every decision with data and KPIs. That habit brought me back to tech.

In September 2026 I completed the MSc in Data Science, Big Data & Business Analytics at Universidad Complutense de Madrid and won the 1st Prize of its scholarship competition. For my thesis I designed and ran a 24/7 capture on a Linux server: 12 million records in 92 days with 99.7% availability, served to a public web app through a REST API. Alongside the MSc I interned at Babel with Azure Data Factory, SQL Server and Power BI. My background in industrial engineering gives me a systems-thinking approach to data architecture.

Experience

Career path

Apr 2026 - Sep 2026

Junior Data Analyst, BI & Azure Data Factory (Internship)

Babel · Madrid (hybrid)

3 to 6 daily data quality issues between SQL Server and Power BI, maintenance and improvement of scheduled Azure Data Factory pipelines, and Power BI reports with DAX. Flagged a redundant employee key inherited from an obsolete legacy system, which led to a development to remove it.
May 2024 - Aug 2025

Health & Safety Manager

ARPO Empresa Constructora · Madrid

20+ workers from up to 4 companies across 4 parallel sites for Canal de Isabel II. Accidents and incidents down 30% through action and communication protocols, every external audit passed, plus internal and stakeholder reporting.
Jun 2022 - Apr 2024

Health & Safety Technician and Supervisor

Ibersys Seguridad y Salud, Nidus · Madrid

Coordination of business activities across industrial and energy projects, structured reports, data validation and document control.
Nov 2017 - Jan 2022

Backend Developer, Financial Systems

BBVA Technology · Madrid

Quality of 3 batch chains with accounting impact (load, processing and reconciliation) in factoring, confirming and revolving, with SQL on DB2, COBOL, JCL and Control-M. Cut the runtime of slow processes by up to 50%. Production incidents and on-demand developments.

Stack and education

What I work with and where I learned it

Technology · Platforms · Languages

Tech stack

Professional use

Azure Data Factory SQL Server Power BI · DAX Python DB2 COBOL JCL Control-M

Projects and training

Databricks Apache Spark (PySpark) Snowflake Parquet Linux · systemd FastAPI MLflow LightGBM scikit-learn LangGraph Anthropic API Gradio Git · GitHub

Education · Certifications · Languages

Education and certifications

MSc in Data Science, Big Data & Business Analytics

Universidad Complutense de Madrid · 2025 - 2026
MSc in Occupational Health & Safety

Universidad Francisco de Vitoria · 2021 - 2022 · Three specializations
BSc in Industrial Mechanical Engineering

University of Huelva · 2011 - 2017
Databricks Accredited Data Governance Fundamentals and 8 Knowledge Badges

Databricks Academy · Oct 2026 · Lakeflow, Spark Declarative Pipelines, Automation Bundles, DevOps, performance and privacy
Databricks Certified Data Engineer Associate

In preparation
Snowflake Platform Skills Badge

Snowflake · Jul 2026
Big Data Algorithms for Engineering

Global Knowledge (SEPE program) · 80 h · Jul 2026
Languages

Spanish (native) · English B2, fluent technical reading and writing

Projects

Featured projects

Published code, documented decisions and measured results.

Data Engineering · Real-Time · ML in production

Delay Prediction for Madrid Commuter Rail

Renfe reports delays only once they are happening and publishes no history. I built the 24/7 capture system that creates that history and the public web app that turns it into an arrival prediction. Master's thesis (UCM), 1st Prize in the scholarship competition, team of five: my part was the infrastructure, data capture, the app and team coordination.

Python GTFS-Realtime AEMET API Parquet systemd · Linux FastAPI Databricks · MLflow LightGBM rclone

12 million records in 92 days, captured every minute, with 99.7% availability
Real-time to timetable match rate raised from 0% to 99.9% by recovering versions the operator does not keep
Immutable raw layer, Parquet analytics layer and daily cloud backup
Content monitoring: alert within 10 minutes if a source stops making sense
LightGBM model: error down from 283.2 to 147.7 s vs historical baseline (−48%)
On real trains: 6.18 min error vs 9.08 for the official timetable

Open the app GitHub

ML · Generative AI · Interfaces

Pump It Up: Water Pump Status Prediction (Tanzania)

Full ML pipeline over 59,000 water pumps in Tanzania: from a stacking ensemble scoring 0.8230 to two usable interfaces: a live individual predictor with Gradio and a conversational assistant powered by the Claude API.

Python scikit-learn XGBoost / LightGBM LangGraph Anthropic API Gradio Pandas SHAP Matplotlib

Overall accuracy misleads: the critical class is 7.3%
2× detection rate of repairable pumps vs. base models
Live individual predictor with per-class probability
Conversational AI assistant over the analysis results

View architecture GitHub

Contact

Connect with me

Phone

+34 692 871 279

Email

ishmaelbovingdon@gmail.com

View profile

GitHub

View profile

Location

Madrid, Spain (on-site or hybrid)

Download CV

PDF