I am David, an AI researcher and Product Builder. My work is about guiding innovation to its ultimate potential.
My philosophy is simple: we must look back to move forward. I use this Sankofa principle to build technology that is both advanced and grounded, acting as a seamless partner for human growth.
Purposeful Innovation in Practice
My research doesn't just live in a lab. They are applied to build real-world products below.
LexLens
SHIPPEDI built LexLens from my Cornell thesis, it's a legal AI that can verify every single source, stopping "hallucinations" before they happen. It helps law students and practitioners build better, faster, and more accurate arguments.
- ✓ Zero-hallucination architecture with transparent source attribution
- ✓ Multi-jurisdictional compliance across 50+ US legal systems
- ✓ Interpretable attention visualization for reasoning chains
- ✓ Constitutional AI framework with bias detection
Tech Stack
Meetviro
SHIPPING JAN 2026As the lead Researcher, I'm helping build an autonomous agent that can attend meetings for you. My research focuses on "social intelligence" teaching Viro to understand human context, know when to speak, act as a seamless trustworthy partner and escalate to a human if things get tricky.
- ✓ Real-time calendar conflict detection and resolution
- ✓ Multi-provider sync (Google Calendar, Outlook, Calendly)
- ✓ Observable SLAs with transparent reliability metrics
- ✓ Transparent human-agent handoff protocols
Ingenium
DELIVERY JAN 2026People drop out of online courses when they get stuck. I'm architecting the AI for Ingenium Academy that spots when a student is struggling and proactively reaches out with encouragement and context-aware help.
- ✓ Hybrid recommendation engine for course discovery
- ✓ Conversational AI tutor with context retention across sessions
- ✓ Proactive student engagement system to prevent dropoff
- ✓ Personalized learning modality adaptation (visual, auditory, kinesthetic)
Goal: It's human-centric design, built to help learners realize their potential.
Current Research
My research began with teaching AI how to reason. My thesis below built a frameworks for verifiable legal reasoning, which I then deployed as the product LexLens.
Currently, my research has evolved into teaching AI how to act. It focuses on the safety and social frameworks for autonomous agents, allowing them to move from simple tools to responsible, "symbiotic partners" in unpredictable human environments.
Research
Search-Augmented Reinforcement Learning for Legal Reasoning
MS Thesis, Cornell 2025 | Advisor: Claire Cardie | Defended September 5, 2025
- • First framework integrating legal database access during RL training episodes (not just inference)
- • Multi-task evaluation with jurisdictional compliance across all US legal systems
- • 10.5pp improvement on LegalBench through tool-assisted training with GRPO
Teaching Experience
CS 2800: Discrete Structures
Graduate Teaching Assistant, Fall 2023 & Fall 2024
CS 1110: Introduction to Python
Graduate Teaching Assistant, Spring 2024
CS 1700: Elements of Artificial Intelligence
Graduate Teaching Assistant, Spring 2025
Let's Connect
If you'd like to discuss research ideas or explore AI safety and reasoning, I'd love to hear from you.
Email: daa238@cornell.edu