Miguel Claramunt

he/him. Research Engineer.

Resume: Miguel Claramunt (.pdf format)

Experience

Research Engineer – Barcelona Supercomputing Center (Nov 2024 – Present)

Machine Translation @ Language Technologies Unit

Private Tutor – Freelance (May 2023 - Present)

Assisted 3 Bootcamp/Master students in the Data Science field totalling +80 hours of tutoring.

Customer Service Engineer – Leeos Merch (Jul 2024 - Sep 2024)

Designed and maintained Customer Service (CS) solutions for 4 online merchandising stores of the largest Spanish-speaking content creators on Youtube and Twitch.

  • Constructed the CS infrastructure from the ground up using Freshdesk, allowing for efficiently resolving tickets during peak season (~5.5k new monthly tickets) while maintaining ~90% Resolution SLA.

  • Upgraded specific routines to improve Average Handling Time (AHT) in 25% and daily CS processes up to 70%, reducing costs by ~6k annually.

Data Analyst – SDG Group España (Sep 2023 - Apr 2024)

Developed data-driven solutions for a leading banking firm, enhancing actuarial and organizational decision-making through cross-functional collaboration.

  • Maintained an internal application serving up to 40k daily users using Qlik Sense, which required near real-time updates for processing ~10k records per hour.

  • Supported clients and products through customized data pipelines, using Oracle Data Integrator and SQL to integrate different data sources with ~500k monthly records.

  • Identified inefficiencies on client’s internal tool, reducing costs by ~20k monthly, leveraging potential contracts and improving consulting firm’s value proposition.

  • Applied Agile development life cycle within two teams of 6 and 8 developers.

Undergraduate Research Assistant – Universitat de València (Apr 2022 - Jul 2022)

Constructed end-to-end applications to tackle fake news classification using stylistic features. This research was part of REMISS project, led by Eurecat.

  • Researched and constructed datasets, evaluating different text typologies to maximise classification performance: ~500k tweets and ~50k news articles.

  • Designed a comprehensive text preprocessing pipeline using scikit-learn, NLTK and spaCy, enabling scalability to process 5x more documents.

  • Evaluated different techniques for document classification: sparse and dense (embedding). Achieved an F1-Score of up to 77% using embeddings.

Education

Master of Science – Intelligent Interactive Systems (2024 - 2026)

Universitat Pompeu Fabra

Bachelor of Science – Data Science (2019 - 2023)

Universitat de València

Bachelor of Science – Mathematical Engineering in Data Science (2022 - 2023)

Universitat Pompeu Fabra