Albert Molins Diez

Worked on the development of scalable backend systems and automation solutions using Python and TypeScript, building APIs with FastAPI and NestJS. Developed and integrated LLM-based applications to automate workflows and support real-world business use cases. Implemented Retrieval-Augmented Generation (RAG) architectures, combining LLMs with vector databases such as Qdrant to enable semantic search and context-aware responses. Performed prompt engineering and iterative prompt optimization, evaluating model outputs and refining prompts through systematic testing and response analysis to improve reliability, accuracy, and performance in production environments.

I worked on the development of high-performance FastAPI-based solutions for an intercontinental submarine internet operator, as well as for an NPL (Non-Performing Loans) management system, using Python as the core technology. For the submarine network operator, I contributed to the implementation of scalable REST APIs migration (Java to Python) to support operational processes, data monitoring, and integration with internal systems. The focus was on performance, reliability, and clean architecture, ensuring that services could handle large volumes of data and mission-critical workflows across geographically distributed infrastructure. In parallel, I developed backend solutions for an NPL management platform, building secure and e icient APIs to manage financial data, business rules, and reporting processes. This included designing database models, implementing business logic, and ensuring data integrity and traceability. I worked closely with stakeholders to translate complex operational and financial requirements into robust technical solutions. Across both projects, I applied best practices in API design, asynchronous programming with FastAPI, structured and modular code organization, testing, and documentation, contributing to maintainable and production-ready systems.

I worked as a Software on the Gavius public sector project, focused on building a virtual assistant for social assistance management. My role involved developing and maintaining backend services in Java to support the processing and management of social aid applications, ensuring reliability, scalability, and integration with existing public administration systems. Over time, I expanded my responsibilities beyond core backend development. I designed and implemented ETL processes using Python to extract, transform, and load data from multiple internal and external sources, improving data consistency and enabling better decision-making. I worked extensively with both structured (SQL) and unstructured (NoSQL) databases, handling data modeling, query optimization, and large administrative datasets. Additionally, I contributed to service development and integrations using TypeScript and Node.js, supporting complementary microservices and APIs within the ecosystem. I also gained hands-on experience with AWS, participating in cloud-based deployments, environment configuration, and the integration of managed services to ensure scalable and resilient infrastructure. Throughout the project, I collaborated in agile, cross functional teams, following best practices in code quality, peer reviews, and continuous integration.