Alberto Penas Díaz

• Fine-tuning large language models with QLoRA for structured output generation, targeting automated Grafana dashboard construction from natural-language clinical descriptions.
• Accepted oral presentation at IWINAC 2026 (Springer LNCS): Estimating MMSE Scores from Conversational Transcripts Using Quantized LLMs.
• Conducted systematic literature reviews on state-of-the-art ML, RL and explainable AI methods to inform the design of clinically grounded AI frameworks.

• Designed and developed hardware-specific firmware frameworks for IoT devices using MQTT, LoRaWAN and NB-IoT protocols across heterogeneous device fleets.
• Collaborated with firmware engineering teams to integrate OTA update mechanisms with versioned image management and automatic rollback on failed flashing cycles.
• Monitored field device telemetry, identified throughput bottlenecks through performance profiling and delivered targeted optimisations that measurably increased message throughput.

• Developed high-performance algorithms in C++ for computational number theory, executed at scale on the HPC-DRAGO Spanish supercomputer.
• Implemented hybrid parallelism (MPI + OpenMP) for both distributed and shared-memory execution models, achieving significant reductions in wall-clock time for large-scale primality and factorisation workloads.
• Designed memory-efficient data structures to handle the volume and precision requirements of extended-range mathematical computations.

• One-on-one tutoring for university students in C and POSIX systems programming, covering memory management, process synchronisation, file-system interfaces and kernel-level debugging.
• Adapted instruction to individual learning pace, with a consistent record of students achieving significant grade improvements.

A fully distributed, end-to-end encrypted messaging system operating over LoRa (EU-868 MHz band) with no reliance on centralised infrastructure or internet connectivity. Built to explore mesh networking and off-grid communication.

A fully functional concurrent Unix shell implemented in C targeting the RISC-V (RV32I) ISA. Validates the intersection of POSIX compliance with RISC-V architecture—cross-compiled and verified through instruction-level emulation.

A hybrid client-server and peer-to-peer file transfer system using TCP/IP sockets for data transfer and ONC-RPC for service registration and peer discovery. Designed to handle concurrent file transfers without data races.

A complete compiler pipeline for Viper, a statically-typed language with Python-inspired syntax. Built from scratch: lexer, LALR(1) parser, semantic analyser and code generator — a full tour of compiler theory put into practice.

Published open-source Python package on PyPI providing interactive visualisation of graph data structures and Binary Search Trees. Originally built as a pedagogical aid; now used by students worldwide.

[1] Estimating MMSE Scores from Conversational Transcripts Using Quantized Large Language Models. Alberto Peñas Díaz, Miguel Ángel Patricio Guisado, José Manuel Molina López, Antonio Berlanga de Jesús IWINAC 2026 — Springer Lecture Notes in Computer Science (LNCS) (2026). Accepted oral presentation.

Universidad Carlos III de Madrid

Example Conference