AI specialist at the University of Alberta and machine-learning consultant at Alberta Health Services. Her work connects clinical applications, model evaluation and infrastructure for deploying AI within a health system.[5][4]
Start with Berta: an open-source, modular tool for AI-enabled clinical documentation →
Education
Links
Website University lab
X · samridhivaid Machine Learning Consultant · Alberta Health Services
AI Specialist, Faculty of Medicine and Dentistry · University of Alberta
Career
Vaid’s experience includes software engineering at Samsung Research and Development, graduate research at the University of Alberta, consulting at Amii and mentoring through AI4Good Lab. She now works on clinical AI systems and their evaluation. Her university lab describes her focus as MLOps and generative AI for healthcare.[4][5]
Contributions
Vaid is the first author of the Berta paper, describing a modular clinical-documentation system. Speech recognition produces a transcript, a language model creates a draft using a configurable note template, and a clinician reviews and edits it before transfer to the health record. The modules can use different models and secure deployment environments. The paper reports an eight-month deployment across 105 urban and rural Alberta facilities involving 198 physicians.[1]
Tools, writing & talks
Writing, research & talks
Start here
The original paper describes the scribe architecture, secure health-system integration and observed deployment.
Start with the architecture and clinical workflow before the deployment figures. Familiarity with speech recognition and language models helps; the paper shows where templates, data boundaries and clinician review enter the system.
First author[1]- Context Engineering in 2026 — Louis-François Bouchard, Omar Solano & Samridhi Vaid, Towards AIyoutube.com
Across 11 context presets measured on an open source AI tutor, sending the full history beat every compaction technique on recall, cost and latency at once, because rewriting context invalidates the prompt cache that serves most of the tokens.
Catalog workOriginal source