Meet the Developer

Award-winning AI Engineer revolutionizing biomedical research with cutting-edge LLM orchestration

Dongyu Lin

🏆 2025 Spark Tank National Science Star Award Winner

AI Engineer and entrepreneur who built two production AI systems at FDA: the PubMed AI Screening Platform and AskFDA RAG System. Sole developer of advanced RAG architecture with LLM orchestration, enabling AI-powered literature screening and private internal document Q&A to accelerate biomedical research. Skilled in RAG, LLMs, LangChain, vector search, and full-stack development

🤖 AI Innovation

Built RAG system for FDA internal document Q&A and AI Agent screening tool to power literature reviews.

🔧 AI Tech Stack

LangChain, Ollama, Chroma DB, RAG Architecture, LLM Orchestration, Prompt Engineering, Agentic AI

📊 Production Impact

83% time reduction for 70+ FDA scientists • 13,300+ articles screened • 92% precision AI responses •20× faster document retrieval

🚀 Entrepreneurship

Self-funded education through $2.2M+ ML-driven e-commerce business • Launched pubmedaiscreen.com AI platform

PubMed AI
Literature Intelligence Platform
Sign In
2025 Spark Tank National Science Star Award Winner

PubMed AI

The Future of Literature Research

Transform your research workflow with our revolutionary AI-powered platform. Trusted by leading research institutions, our system has already helped scientists save thousands of hours while screening massive volumes of biomedical publications.

13,300+
Articles Analyzed
83%
Time Reduction
1,100+
Hours Saved
80%
Less Hallucinations

Proven Results That Speak for Themselves

Built from scratch, delivering exceptional performance at scale

83%
Faster Literature Reviews

Cutting manual review time from days to hours

13,300+
Articles Processed

High-volume screening with AI precision

1,100+
Research Hours Saved

More time for breakthrough discoveries

Enterprise-Grade Security | JWT Authentication | Distributed GPU Processing

Powerful Research Tools

Everything you need to conduct efficient and thorough literature reviews, powered by cutting-edge artificial intelligence

Smart Literature Search

Self-validating AI agents generate precise PubMed queries with 80% fewer hallucinations through advanced prompt engineering.

AI-Powered Analysis

Full-stack LLM orchestration with React/Next.js frontend and Python/FastAPI backend, featuring distributed GPU processing.

Advanced Analytics

MongoDB-powered analytics with automated Excel reporting via Pandas and real-time insights dashboards.

Lightning Fast

Production-scale pipeline with Docker/nginx deployment, priority queuing, and intelligent failover mechanisms.

How It Works

Three simple steps to revolutionize your research process

1

Self-Validating AI Search

Our proprietary AI chatbot generates PubMed queries with 80% fewer hallucinations through advanced prompt engineering and validation.

2

LLM-Powered Screening

Full-stack AI orchestration with distributed GPU load balancing screens thousands of articles using state-of-the-art language models.

3

Automated Insights

Generate comprehensive Excel reports with Pandas, complete with priority queuing and failover mechanisms for reliability.

Ready to Transform Your Research?

Join research teams worldwide who are using PubMed AI to accelerate their discoveries and save thousands of hours on literature reviews