Passionate about leveraging artificial intelligence to build scalable, innovative solutions that make a real-world impact. (And I've done so already 😉)
- Leading a team at Capital One to implement an AI Powered Slack solution for a fraction of the cost!
- Mentoring engineers to develop and implement Enterprise-Friendly AI/ML Guardrails to mitigate bias & hallucination.
- Preparing for the AWS Machine Learning Engineer Associate exam (wish me luck!).
- Designing an enterprise-scale LLM course—from ingestion to deployment. Interested? Let’s chat!
-
🏫 Education: Graduated from Duke University with a degree in Electrical & Computer Engineering and Computer Science.
-
🌱 Currently Exploring: Deep diving into Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and AI safety protocols.
-
💼 Professional Experience:
- Capital One:
- Collaborated with over 50 cross-functional stakeholders, solo-driving a $300M+ proposal with buy-in from executive leadership.
- Optimized real-time ML fraud detection systems, reducing latency by 70% using ElastiCache/Redis.
- Spearheaded the integration of vector databases for low-latency access in distributed systems.
- Managed LLM projects with guardrailed LLaMa 3.1, achieving 90% accuracy and reducing bias by 37%.
- Led a team to victory in the Capital One Enterprise-Wide Hackathon, developing a RAG-powered Slack/VSCode agent that boosted developer productivity by 80% and impacted 30,000+ employees.
- Natter:
- Architected high-resilience, event-driven distributed systems for 10,000+ concurrent users.
- Developed fine-tuned, open-source LLMs utilizing zero-shot learning and novel embedding techniques.
- Increased user engagement by 40% through AI-driven user matching and predictive analytics.
- Tuned Systems Inc.:
- Created a high-frequency trading alert system with low-latency routing, achieving an 85% success rate in client-facing trades.
- Implemented optimization algorithms and trade execution logic using TypeScript, GraphQL, and SQL.
- Capital One:
-
🔭 Current Projects:
- AI Celebrity Replica:
- Fine-tuned Phi-3 and Mistral 8B models for celebrity emulation.
- Improved response accuracy by 40% using LangFuse, PostgreSQL, and Tavily.
- Orchestrated AWS services for 99.9% uptime, enabling retrieval-augmented generation.
- Exploring AI Ethics and Safety:
- Focused on implementing ethical considerations and safety measures in deploying large-scale AI systems.
- AI Celebrity Replica:
-
🏆 Achievements:
- AWS Certified Solutions Architect Associate
- HackDuke 2021 Health Sector Winner
- Capital One Innovation Day Winner
- AWS InCommunities Scholar
-
💬 Ask Me About: LLMs, Distributed Systems, Cloud Architecture, AI Ethics, and Machine Learning Engineering.
-
📫 How to Reach Me: [email protected]
-
⚡ Fun Fact: I play seven different instruments!
Feel free to reach out if you're interested in collaborating on AI projects or just want to chat about the latest advancements in technology!