Yassir Dembri

About Yassir Dembri

Biomedical Engineering student at Boston University with various HealthOps experience. I’m interested in how AI can actually improve healthcare in real, regulated environments; where validation, risk, documentation, and workflow fit matter as much as model performance. I\'m currently expanding into machine learning, data-driven modeling, and AI systems, through my concurrent gaduate program at the University of Colorado (Boulder) with the goal of applying these technologies to improve healthcare, medical devices, and biomedical research.On LinkedIn, I share- Practical breakdowns of clinical AI papers (what’s useful vs. what’s hype)- Mini-projects and demos in Python (evaluation, calibration, drift, monitoring)- Medtech-minded lessons on verification/validation, failure modes, and safetyMy goal is to be a clear, reliable voice at the intersection of AI + MedTech + quality.If you work on clinical ML, SaMD, medtech product, quality, or validation, I’d love to connect and learn from your work.

Experience

1. Trainee

   Ai Training

   Feb 2026 — Present · Кембридж, MA, US

   8-week BootcampAnalyzed core AI concepts including transformer architectures, embeddings, and LLM systemsExplored and experimented with retrieval-augmented generation (RAG), including chunking, vector embeddings, and semantic searchWorked with and evaluated LLM outputs, identifying failure modes such as hallucination and inconsistencyInvestigated agent-based workflows involving tool use and multi-step reasoningExamined real-world challenges including model drift, reliability, and deployment constraintsApplied learned concepts to build personal projects using an LLM-based system with structured outputsExperimented with prompt and context engineering to improve output accuracy and grounding

Education

• University of Colorado Boulder

  Master's Degree, Artificial Intelligence

  2026 — 2028

• Boston University

  Bachelor of Science - BS, Biomedical/Medical Engineering

  2022 — 2027