Ghodrat Aalipour - Homepage
Gordi Aalipour | Principal Applied AI Scientist

About Me

I design and build intelligent AI systems that combine LLMs, knowledge bases, and retrieval to solve complex problems in dialogue, search, and reasoning.

Industry Experience:

• Gusto – Principal Applied Scientist
  Agentic AI Systems for legal & financial tech, task-oriented dialogue, disambiguation agents
• Securiti.ai – Lead Data & AI Scientist
  RAG systems, multi-modal QA, document intelligence, LLM safety & security
• Amazon Alexa AI – Applied Scientist
  Retrieval & ranking for Knowledge Graph Question Answering, temporal reasoning
• Walmart Global Tech – Staff/Senior Data Scientist
  Conversational AI, Knowledge Graphs, NLU models for shopping assistants

Academic Experience:

• Adjunct Professor – Rochester Institute of Technology
• Visiting Researcher and Lecturer – University of Colorado Denver

Education

Ph.D. & M.S. in Mathematics
Sharif University of Technology, 2013
My doctoral research was at the intersection of Linear Algebra and Discrete Mathematics, focusing on Spectral Graph Theory and its applications to theoretical computer science.

M.S. in Computer Science
Rochester Institute of Technology, 2018
Focus on Natural Language Processing and Computer Vision.

My journey from mathematics to AI has been fascinating—I became interested in machine learning through the Laplacian Matrix in Spectral Clustering. Today, I see the same mathematical concepts powering Graph Convolutions and Graph Neural Networks.

Research & Career Interests

My current focus is on building Agentic AI Systems—designing task-oriented dialogue agents that can reason, plan, maintain memory, and interact with tools to solve complex, multi-step problems. I'm particularly interested in compound AI systems that combine LLMs with knowledge bases, retrieval systems, and specialized sub-agents.

I have extensive experience with Retrieval-Augmented Generation (RAG) systems, working on everything from query/corpus expansion to multi-index hybrid search strategies. My work spans document intelligence (VQA, TableQA, PDF parsing), semantic search, and building enterprise-scale retrieval systems.

I'm also passionate about Knowledge Graphs and their integration with LLMs—including link prediction, entity disambiguation, knowledge-based question answering, and graph neural networks. My background in mathematics gives me a unique perspective on the theoretical foundations underlying modern AI systems.