raul escandón

sep 2026 to now · education · London, UK

MSc Artificial Intelligence for Biomedicine & Healthcare

University College London (UCL)

nov 2024 to sep 2026 · work

AI Engineer

Arkangel AI

• Mid AI Engineer (jan 2026 to sep 2026)
  • Built a multi-agent system that extracts diagnoses, insurance and audit data from medical records.
  • It audits 1,000+ medical bills a day, cutting labor cost by 60% and doubling manual review coverage.
  • Co-authored four clinical AI preprints: SOAT claims, billing review, clinical coding and sepsis severity.
• Junior AI Engineer (may 2025 to dec 2025)
  • Found the Colombian PUMA extension had no training data, so I built a PANDORA-based LLM extraction system on AWS to pull records at scale.
  • Processed 2,000+ files in parallel at 20 seconds each, turning messy records into usable training data.
• AI Intern (nov 2024 to apr 2025)
  • Joined the pre-training data team for a Colombian extension of PUMA Score, built with a local hospital and AstraZeneca.
  • Prepared the clinical dataset for pre-training with statistical analysis, coherence checks and imputation.

apr 2024 to oct 2024 · work · Lima, Peru

Software Engineer

Indigital XP · Intercorp Retail (Promart)

• Improved Core Web Vitals by 30% by rebuilding Promart's e-commerce pages and replacing legacy jQuery and SCSS.
• Shipped the redesigned homepage from the design handoff.
• Built a Google Cloud service that flagged anomalous sale prices and hid them until a human approved.

mar 2017 to aug 2022 · education · Lima, Peru

BSc Computer Engineering

Universidad de Ingeniería y Tecnología (UTEC)

• Graduated in the top tenth percentile and made the Dean's List for academic excellence and research (2020).
• Published five papers as an undergraduate: two international (one journal, one conference) and three national conference papers.
• Selected for the Eastern European Machine Learning Summer School (EEML) in 2021.

• Robotics & Mechatronics Lab (jan 2022 to dec 2022)

  UTEC · advisor: Deyby Huamanchahua

  • Built Amaru Maki, a wearable that measures elbow and wrist range of motion for physiotherapy.
  • Rebuilt it as Amaru Maki v2 with four IMUs, 60 Hz wireless streaming and quaternion-based joint tracking.
  • Co-advised undergraduate theses on exoskeleton joint measurement.
• Research Center for Heritage Conservation (may 2021 to dec 2021)

  UTEC · advisor: Jimmy Tarrillo

  • Built two environmental monitoring stations and the protocol linking an Arduino to a PSoC 5LP.
  • Designed real-time humidity and temperature dashboards used to assess conservation conditions.
• Systems & Signal Processing Laboratory (apr 2020 to dec 2020)

  UTEC · advisor: Christian Flores

  • Designed a data-acquisition protocol and a Fourier decomposition method that separated muscle fatigue from non-fatigue states.
  • Classified two-, three- and four-class hand movements from sEMG signals using fractal features, self-organizing maps and SVMs.
  • Turned the work into two papers: grasping classification at the Brazilian Technology Symposium and sEMG clustering at IEEE EIRCON.
• Robotics & Mechatronics Lab (sep 2019 to feb 2020)

  UTEC & Purdue University · advisor: Ruth Canahuire

  • Built an autonomous beach-cleaning robot with sensor fusion and a master–slave motor protocol.
  • Coordinated the Peruvian team's assignments and deadlines using agile methods and OKRs.
• Robotics & Mechatronics Lab (aug 2017 to aug 2019)

  UTEC · advisor: Ruth Canahuire & Joseph Campion

  • Built a gecko-inspired climbing robot and designed its active suction-cup mechanism.
  • Implemented task-space kinematic control for a delta robot and a minimax checkers opponent.
  • Published a first pair of conference papers: the climbing robot at INTERCON 2019 and checkers at INTERCON 2018.