Toronto, Canada
PhD in Neuroscience · AI/ML & Data Science Specialist · Technical Product Manager
I take AI products from research to production — from PHIPA-compliant clinical Generative AI platforms to enterprise ML systems delivering $2M in annual savings.
I'm Amir Zarie, a Toronto-based technologist with a PhD in Neuroscience from York University. I work at the intersection of AI engineering and product leadership — owning the full lifecycle of AI initiatives from prototype through compliance, launch, and scale.
As co-founder of PocketReport AI, I took a PHIPA-compliant Generative AI clinical reporting platform from prototype to production across 10 clinics, saving each practitioner roughly 8 hours of administrative work per week. At Telus, I architected an RNN/ARIMA anomaly detection system and re-engineered enterprise data pipelines that delivered $2M in annual savings.
Whether building RAG and agentic LLM architectures or translating between engineering teams and executives, I turn research-grade ideas into products that ship.
Coach cross-functional consultant pods through Agile, sprint-based delivery and hands-on AI/ML labs — designing RAG and agentic LLM reference implementations (LangChain, LangGraph, CrewAI) and tracking readiness KPIs with Power BI and Tableau dashboards.
Owned the full lifecycle of a PHIPA-compliant Generative AI clinical reporting platform (Python, FastAPI, LangChain, FAISS) — secured $30k in early funding and scaled from prototype to production across 10 clinics, saving each practitioner ~8 hours of administrative work per week.
Launched a data consultancy generating $35k in revenue — built custom predictive models and analytics pipelines for healthcare, fintech, and education clients, managing every engagement end to end.
Led a multi-modal ML diagnostic pipeline (EEG, behavioral, genomic) to 90%+ precision, commercialized it as a React/FastAPI SaaS platform, and cut annual OpEx by $42k through GCP and PostgreSQL optimization.
Architected a high-throughput Python/SQL anomaly detection system (RNN + ARIMA) for enterprise-scale real-time auditing, and delivered $2M in annual savings by re-engineering data pipelines and Looker dashboards.
Full-stack Retrieval-Augmented Generation platform that aggregates, embeds, and indexes Canadian federal legislation, parliamentary debates, and proposed regulations.
AI-based news aggregator featuring conversational search with interactive sentiment-analysis models and visualization pipelines.
Secure, high-performance logistics platform — migrated to an Nginx infrastructure with auto-renewed SSL and optimized real-time API retrieval with smart caching.