I see INCIDENTS: an LLM that flags incidents in real customer cases.
Eleven projects. Six in production at Telefónica, two theses and three of my own.
I work where generative AI stops being a demo and enters a real process, with real users, real latency, real cost and real adoption.
The code is half the job. The other half is picking the process worth automating, measuring it, and getting the team to actually use it.
I'm Juan Carlos Jiménez, an AI Engineer at Telefónica / Movistar Plus in Madrid, Spain. For three years I have designed, shipped and run generative AI in production: LLM systems, RAG, fine-tuning, agents and their evaluation, on top of a background in computer vision. I hold a BSc in Computer Engineering from Universidad Europea and I am studying an MSc in Artificial Intelligence at UNIR.
Telefónica / Movistar Plus · Tres Cantos, Madrid
LLM solutions for digital operations and customer care: incident detection, a generative chatbot, fine-tuned NLP, review automation and a self-service portal. Also the link between business, operations and development, prioritising what to automate.
UNIR
Universidad Europea
MongoDB
MongoDB
Open to conversations about LLM systems that have to work outside a demo.