“Signal Detection at SFDA: A Comprehensive Analysis of Current Practices and Future Directions”
Journal of Evaluation in Clinical Practice · 31(3):e70062 · First author
AI & Healthcare Architect
AI Regulatory & Healthcare Tech Consultant. I help organizations build compliant AI systems for Healthcare. Expert in LLMs, Drug Regulatory, and Drug Safety.
Focus
AI in Pharmacovigilance & Regulatory Tech
Shipped projects
+9
Drug Regulatory | AI
Safety by Design
About
I wear two hats at SFDA: Director of the Medication Error Department, where I lead the national effort to prevent look-alike drug packaging from harming patients, and Head of the AI Team, Drug Sector, where I drive digital transformation with a 15+ person multidisciplinary team.
My path is uncommon: PharmD from MCPHS Boston, a Master's in Applied AI, and now pursuing a PhD in Computing. I don't just advise on AI in healthcare—I build the production systems regulators use daily.
This blend earned me the SDAIA & Stanford Hackathon Championship and seats on the ICMRA AI Steering Committee and as Chair of ISoP's AI in Pharmacovigilance SIG.
My mission? Make medications safer—one algorithm at a time.
Highlights
0
AI Tools Shipped
0
Publications
0
Years Experience
📄 Selected Research
Journal of Evaluation in Clinical Practice · 31(3):e70062 · First author
Guiding organizations through the adoption of AI in regulated healthcare environments.
Designing scalable, compliant regulatory tech stacks.
From concept to deployment: building RAG pipelines, agents, and predictive models for health data.
Trusted stack
Proven Track Record
Automated prioritization of safety signals, reducing manual triage time by 40%.
Flask · Gemini · Data Viz
Accelerated benchmarking of regulatory documents against gold standards.
GPT-4 · PDF · Benchmarking
Instant similarity search across thousands of SPC documents.
FAISS · TF-IDF · SPC
Streamlined signal detection with multi-agent analysis and visualization.
Flask · LangChain · OpenAI
Intelligent support for QPPVs with historical context retrieval.
GPT-4 · FAISS · Hybrid
AI-powered detection of look-alike medication packaging errors.
Vision · GPT-4.1-mini · Dropzone
Real-time intelligence dashboard for monitoring regulatory updates.
FastAPI · SQLite
Citation-backed regulatory intelligence for confident decision making.
Flask · OpenAI · RAG
Automated comparison of product characteristics to ensure consistency.
Flask · TF-IDF · PDF
Automated prioritization of safety signals, reducing manual triage time by 40%.
Launch ↗Streamlined signal detection with multi-agent analysis and visualization.
Launch ↗Citation-backed regulatory intelligence for confident decision making.
Launch ↗Accelerated benchmarking of regulatory documents against gold standards.
Launch ↗AI-Powered Medication Safety
1,500+
Packaging errors annually
Patients · Regulators · Pharma
All at risk
2-3 hours
Manual review per package
Dual-Engine AI
Hashing + Vision LLM
Upload → Analyze → Report
30 seconds
500+ scanned
47 flagged · 40% faster
Let's Talk Strategy
Schedule a focused conversation about your AI and regulatory challenges. No sales pitch—just strategic insights tailored to your needs.
Expert Analysis
Get actionable insights on your regulatory AI challenges
Custom Roadmap
Walk away with a clear next-step strategy
Zero Commitment
Free consultation with no obligations
📅 Select a time below that works for you. I'll send a confirmation within 24 hours.
Script Doctor
Paste an excerpt of a regulatory document — safety report, labelling text, procedure — and the Script Doctor will analyse it for regulatory implications and suggest next steps.
Do not paste confidential, patient-identifiable, or commercially sensitive text. Excerpts are sent to OpenAI for processing and leave this server. This site does not intentionally retain them; OpenAI handles them under its own data-retention terms.
Analyses are for informational purposes only and do not constitute legal or regulatory advice.
Contact
Prefer email or a quick call. I respond within 24 hours.