Andrés Fernández

aka Jorge, Jorge Andrés, Fernández Jiménez, or just fernandrez

Machine Learning Engineer & AI Architect

15+ years building enterprise AI systems across 4 continents — from actuarial pricing models in Medellín to multi-agent platforms in London. Currently leading a confidential structured-communications automation program through iSeeCI and Machine Learning Engineer at John Snow Labs, founder of iSeeCI and Tailoredia.

What I do

I design and ship production AI systems built on Large Language Models, Retrieval Augmented Generation (RAG and GraphRAG), and Knowledge Graphs. My work spans NLP, document intelligence, and multi-agent orchestration. I lead with Claude Code and Cursor for AI-assisted engineering.

Recent impact

  • 40% inference time reduction on a carbon-emissions ML model at Mondra (London) while preserving prediction accuracy.
  • End-to-end RAG system for personalized meeting summaries at Combine AI (Eric.ai project), built on open-source LLMs (Mistral, Llama) with async/await + Celery on Redis.
  • Since April 2026, an undisclosed engagement through iSeeCI with a major structured communications automation firm — agentic document intelligence and case-workflow automation in production. Client confidential under NDA.
  • Led Timmy, Teramind's Workforce Intelligence Copilot (until March 2026) — natural-language access to security and productivity data for CISOs and HR leaders.
  • Production NLP pipelines on Apache Spark at John Snow Labs since 2019, processing terabytes of documents into relation graphs with SparkNLP.

Why it matters

Enterprise AI fails when the gap between research and production is underestimated. According to a 2023 Gartner poll, only ~10% of generative-AI projects reach production scale. I bridge that gap by combining solid MLOps foundations (Docker, Kubernetes, multi-cloud) with rigorous evaluation harnesses, grounded retrieval (GraphRAG over flat RAG for multi-hop reasoning), and an unwavering focus on the actual business question.

Background

MSc in Software Engineering and AI from Universidad de Málaga (2012, robotics-focused thesis on sound-source localization with neural networks), Actuarial Science from Universidad Nacional de Colombia, and Mechatronic Engineering from EIA (thesis: neural-network actuarial platform). Coursera Deep Learning Specialization (deeplearning.ai). Trilingual: English, Spanish, Italian.

If you want to ship AI that works in production, get in touch at fernandrez@iseeci.com or scroll to the contact form.

Ask Fernandrez