Carlos Barranquero

I was a Technical Lead specializing in Intelligent Document Processing (IDP).

We used AWS to create an email automation platform in the insurance sector. We used NLP and Machine Learning models to classify and extract specific data from claims and policies.

We processed around 3,000 emails daily with 96% accuracy, reducing manual effort (HITL) for Mapfre (USA) and Verti (Germany).

I was part of an international team, and my responsibility was to automatically extract data from invoices, medical documents, and contracts using AI models.

Over time, I was given the responsibility to create a generic extraction model to attract new clients, perform quick demos with their data, and showcase the product's capabilities.

I was assigned a team of 4 people with no prior AI experience, trained them remotely and in English. We acquired 2 new clients and expanded the product to the USA.

I joined this startup without knowing how to code. The company invested heavily in me, training me from scratch in programming and artificial intelligence.

In a short time, I was given significant responsibility. I was in charge of several AI applied projects in engineering and business intelligence. Some of the things I did:

- Water turbidity prediction system (Acciona Qatar): I developed an AI application for the Ras Abu Fontas desalination plant. The system predicted water turbidity levels, allowing the operations team to anticipate the best time to shut down the plant. It was in production for several years.

- Automatic license plate detection system: For a Belgian client, I designed and implemented a computer vision system capable of identifying and reading license plates from cars in multiple parking lots, automating access control.