MB.

Madrid, Spain

Miguel Bravo

AI & Data Engineer

I build useful systems where data, machine learning and software meet.

LinkedIn ↗︎ GitHub ↗︎ Résumé 2026 ↗︎

My work starts with messy real-world data and ends with something people can understand, trust and use.

Current focus

Process intelligence, applied AI and measurable enterprise value at Celonis.

Published

Two peer-reviewed papers on AI-driven heat-risk systems, in Elsevier journals.

 Explore the work

01About

I’m an AI & Data Engineer from Granada, now based in Madrid. I work across the full path, from models, APIs, design and launch of AI agents and orchestrated solutions all the way to explaining it to executives so they can understand it and act on it.

My background in robotics taught me to think in systems. Research in climate informatics taught me to make those systems rigorous. Consulting keeps the result grounded in what people and organisations actually need.

  • Executive communication

    Presenting technical solutions and value cases to C-level stakeholders.

  • Customer-facing

    Working directly with enterprise customers to turn process data into measurable outcomes.

  • On stage

    Talks and papers at WUSS, MWSUG and the RVAtech Data + AI Summit.

Miguel Bravo
Granada → Málaga → Ohio → San Diego → Madrid Engineer, researcher, compulsive builder.
  • Universidad de Málaga
  • Miami University
  • San Diego State University
  • SDSU Research Foundation
  • Premier Analytics Consulting
  • Celonis

02Experience

Résumé Updated November 2026 · PDF · 1 page View the full résumé

Mar 2026 — Present

Applied Value Engineer

Celonis

  • Diagnosing operational inefficiencies in the end-to-end processes of large enterprises from process data, and quantifying their financial impact as measurable value and ROI.
  • Designing and prototyping AI agents and orchestrated automations on Celonis’ process intelligence and context layer, in domains such as supply chain.
  • Presenting solutions and value cases directly to customers and C-level stakeholders, translating complex AI and data concepts into clear business outcomes.
  • Process Intelligence
  • AI Agents
  • Value Engineering

May 2025 — Feb 2026

AI & Data Science Consultant

Premier Analytics Consulting

  • Designed data pipelines and FastAPI backends to automate ingestion and transformation of large research and environmental datasets across AWS S3 and on-premise servers.
  • Built and deployed containerised AI systems with Docker and Ubuntu, improving reproducibility and deployment efficiency.
  • Developed a localised multilingual assistant for summarisation, translation and retrieval using open-source LLMs (Mistral-7B, LLaMA 3), tuning latency and cross-lingual accuracy.
  • Collaborated on end-to-end product delivery, integrating ML components with React frontends and REST APIs.
  • Python
  • FastAPI
  • Docker
  • LLMs

Jun 2024 — Feb 2026

AI & Data Scientist

iHeatApp · SDSU Climate Informatics Lab

  • Built a bilingual AI-driven product for agricultural communities that maps real-time WBGT heat risk and gives contextual safety guidance through a custom LLM chatbot.
  • Built and validated hourly forecasting pipelines on NOAA NBM data (Herbie/xarray), with geospatial rendering for the Imperial Valley.
  • First author of the paper on the system architecture and modelling framework (IJDRR, Elsevier) and co-author of a second paper on the AI chatbot (Climate Services).
  • Python
  • xarray
  • Geospatial
  • AI

Aug 2023 — May 2024

Data Science Research Assistant

Center for Information Convergence and Strategy

  • Applied multivariate statistics and time-series analysis to nutrient dynamics and eutrophication in Irish agricultural catchments.
  • Developed automated R pipelines for cleaning, interpolation and quality-flagging, improving reproducibility.
  • Supported predictive modelling of nutrient trends with regression and spatial-temporal clustering.
  • R
  • Research
  • Spatiotemporal data

03Publications

PublishedFirst author

iHeatApp: A climate data visualization and delivery technology for mitigating heat exposure and risks in agriculture

First-authored paper on the architecture and modelling framework behind iHeatApp, the bilingual heat-risk platform for agricultural communities.

PublishedSecond author

An AI visual chatbot system for climate heat risk communication and decision-making in at-risk agricultural communities

Co-authored paper on the AI visual chatbot that communicates heat risk and supports decisions in at-risk agricultural communities.

Journal metrics from ScienceDirect (November 2026). Citation counts update live from OpenAlex.

04Projects

Selected work

Three flagship projects first, then nine more, roughly ordered by how well they show how I think and build.

Editorial heat-risk map of Imperial Valley
P012024—26

iHeatApp

A bilingual platform that turns hourly weather forecasts into local heat-risk maps and contextual guidance.

  • Climate AI
  • Python
  • EN / ES
Editorial illustration of weather data flowing into compact computing hardware
P022025

Weather with Cloudya

Real-time NOAA climate data, interactive maps and a vision assistant served from a Raspberry Pi.

  • FastAPI
  • React
  • Docker
Editorial illustration of flight paths and a diverging delay signal
P032024

Flight Delay Prediction

Three million flight records, weather context and an honest comparison of seven model families.

  • XGBoost
  • Streamlit
  • WUSS

+More projects

Editorial map of California traffic and demographic patterns
2023

California Traffic Trends

Ten years of traffic, population and income combined in one geospatial analytical story.

  • Tableau
  • R
  • ArcGIS
Editorial illustration of diamond facets and regression surfaces
2024

Diamond Price Modelling

Six regression approaches compared across 53,940 diamonds and transformed targets.

  • Regression
  • Random Forest
  • EDA
Editorial illustration of a relational pharmacy database
2024

PharmaSys

Relational database design, AWS deployment and analytical views for pharmacy operations.

  • MySQL
  • AWS RDS
  • Plotly
Editorial illustration of seasonal retail sales forecasting
2024

Walmart Sales Forecasting

Seasonal weekly sales forecasting for one store through a focused SARIMA comparison.

  • SARIMA
  • Time series
  • R
Editorial illustration of overlapping multivariate observations
2024

Mental Health & Technology

A multivariate study where weak predictive signal became the important methodological result.

  • PCA
  • ANOVA
  • Clustering
Editorial illustration of motion and collision in an interactive game
2023

Getting Started with p5.js

A small interactive game for learning the drawing loop, events and collision logic.

  • p5.js
  • Interaction
Editorial illustration of a searchable university campus map
2023

Interactive Campus Map

A zoomable and searchable campus map with building-level information and imagery.

  • p5.js
  • Maps
  • Search
Editorial world map connecting news stories to locations
2023

News Geovisualisation

Recent news search, geocoded results and the geographic context behind each story.

  • p5.js
  • API
  • Geocoding
Editorial map of changing US crime-rate centroids
2023

US Crime Trend Explorer

Multivariate and multitemporal views of state crime rates and moving geographic centroids.

  • p5.js
  • Time
  • Choropleth

+Beyond the code

Speaking & writing Explaining models, open source and better data workflows.
Miguel Bravo at a data and AI conference

05Contact

Contact & profiles

Based in Madrid, Spain

Email miguelangelbravo2000@gmail.com LinkedIn miguel-angel-bravo GitHub @miguebmv Résumé Updated November 2026 · PDF
Preview of Miguel Bravo's one-page résumé