Shan Rizvi
Context Architect for Artificial Intelligence
Transforming complex workflows through ontology-powered systems and advanced context engineering for AI applications that deliver tangible value
About Shan
I am a context architect, product leader, and technology strategist specialized in launching innovative startups leveraging advanced generative AI, knowledge graphs, and specialized ontologies to automate and enhance complex workflows.
My Recent Projects
- •Ontology-powered graph retrieval to significantly enhance generative model performance and context-awareness.
- •Advanced context engineering for empathic health applications, enabling detailed psychological profiling and precision medical analyses.
- •Deploying Model Context Protocol (MCP) servers integrated with Neo4j databases, enabling generative models to reason using structured psychological and contextual data.
Previously, I built and scaled startups like Just Ads, enabling global brands such as Alexander Wang to optimize multi-platform advertising through intuitive AI-driven automation tools, and Pencil News, a pioneering NLP-driven news app for underserved linguistic markets.
I excel at translating cutting-edge AI research into practical applications that deliver tangible value.
Featured Projects
From emotionally intelligent AIs to advanced reasoning engines, Shan's latest projects showcase his drive to push the boundaries of what AI can do.
Elys is an AI companion designed to offer empathetic conversation, psychological insight, and advice. It uses generative AI combined with psychological profiling to truly understand the user.
- •Understands and adapts to user's personality
- •Offers empathetic conversations and mental well-being support
- •Profiles the user's relationships to give tailored advice
A Model Context Protocol (MCP) server that gives AI systems a form of long-term memory and reasoning ability by storing knowledge in a graph database. This enables large language models to build and maintain a sophisticated knowledge graph in Neo4j.
- •Graph-based memory storage for AI (built on Neo4j)
- •Captures contextual relationships and even emotional nuances
- •Transforms AI assistants from stateless Q&A bots into agents with long-term memory
An AI system that can perform advanced legal reasoning by retrieving information from a law-focused knowledge graph. This approach involves encoding legal knowledge into a graph structure and hooking it up to a generative model.
- •Integrates a knowledge graph of legal precedents into AI's reasoning process
- •Enables the AI to retrieve and cite relevant cases/statutes during analysis
- •Aims for transparent, thorough legal reasoning (useful for lawyers and researchers)
Previous Startups & Achievements
Co-founded this advertising technology SaaS which automated online ad campaigns for global brands. Just Ads enabled companies to manage creative assets, target audiences, budgets, and KPI goals across multiple ad platforms from one dashboard.
Key Achievements
- •Led product strategy and development
- •Served major clients like Alexander Wang, Flightmate, and Lyko
- •Created an easy-to-use performance marketing solution
Co-founded and led one of the first AI-powered news apps for the Urdu language. Shan's team developed groundbreaking NLP tools – the first ever tokenizer, transliteration system, sentiment analyzer, entity recognizer, and recommendation algorithms specifically for Urdu.
Key Achievements
- •Over 30,000 monthly active users
- •300,000+ installs with a 4.7★ rating on Google Play Store
- •Pioneered NLP tools for an underserved language
Let's Connect
Interested in collaborating or learning more about my work?
Reach out at shan@rizvi.nu for collaboration opportunities
© 2025 Shan Rizvi. All rights reserved.