About
Building AI systems that work in the real world, from on-device LLMs in Africa to real-time campus intelligence systems that answer student questions daily.
My Journey
At 20, I've already worked at some of the world's most influential tech companies, building production ML systems that serve millions of users. My work spans the intersection of hardware and software, with a focus on bringing cutting-edge AI to resource-constrained environments.
From developing telemetry-based PyTorch pipelines at Apple to building Llama-powered recommendation systems at Meta, I've consistently pushed the boundaries of what's possible with edge AI and real-world deployment.
Currently, I'm focused on democratizing AI infrastructure through TryTools, while researching bias detection in NLP systems at Howard University. My goal is to make AI accessible and beneficial for everyone, everywhere.
Awards & Recognition
Location
Washington, DC | Howard University
Based in the heart of the nation's capital, studying at a prestigious HBCU.
Education
Electrical Engineering & Computer Science
Major and minor focusing on the intersection of hardware and software systems.
Recognition
National Science & Technology Medals Foundation Award Winner
Recognized for outstanding contributions to AI and technology innovation.
Focus
Edge AI & Real-world Systems
Building AI that works where it matters most - in resource constrained environments.
Technical Expertise
Full-stack capabilities across the entire AI development lifecycle, from research to production deployment.
Languages
AI/ML
Frontend
Backend
Infrastructure
Databases
Experience
Building production ML systems that serve millions of users across leading technology companies and research institutions.
Apple
Software Engineering Intern - Wireless Technologies & Ecosystems
Building intelligent carrier configuration systems for iOS ecosystem. Architected ML pipeline using telemetry data to predict optimal settings across iPhone, iPad, and Apple Watch, dramatically reducing manual triage overhead.
Key Achievements
- Architected production-ready PyTorch pipeline achieving 0.91 AUC on carrier prediction
- Pioneered Bayesian hyperparameter optimization with RLHF-inspired reward mechanisms
- Seamlessly integrated ML models into Apple's CI/CD infrastructure with automated validation
- Accelerated deployment cycle to <10 iterations using telemetry-driven optimization
Technologies
Meta
MetaU Software Engineering Intern
Built full-stack web application with Llama-based chatbot, OAuth-secured flows, and 8 app features. Improved recommendation algorithm by 27% using cosine similarity.
Key Achievements
- Built Llama-based chatbot with context injection and OAuth-secured flows
- Improved recommendation algorithm by 27% using cosine similarity on multi-dimensional vectors
- Built scheduled job (Node-cron) to compute trending artists every 3h using hash maps
- Optimized Prisma/PostgreSQL with indexing + JOIN tuning, reducing query latency 1.5x
Technologies
Howard University
ML/AI Research Assistant - NLP for Bias Detection
Built TF-IDF + C++ tokenizer for microaggression detection achieving 85% F1 score on 10K+ labeled samples in bias-sensitive NLP pipeline.
Key Achievements
- Achieved 85% F1 score on microaggression detection with 10K+ labeled samples
- Built custom C++ tokenizer for efficient text processing and feature extraction
- Reduced feature extraction time by 30% via custom vectorization scripts
- Presented results at Howard University's Undergraduate Research Month with recognition
Technologies
Interested in working together or learning more about my experience?
Projects
Revolutionary AI systems that push the boundaries of what's possible in real-world environments. From edge computing to production platforms.
Showing 13 of 13 projects
AI-powered decision intelligence platform that transforms team decisions into searchable, actionable knowledge. Captures context, reasoning, and outcomes of every decision to build organizational memory.
Key Features
Metrics
Tech Stack
AI-native backend platform that lets founders deploy real infrastructure in seconds from a single prompt. Auth, payments, notifications, analytics, and ML pipelines all live and production-ready.
Key Features
Metrics
Tech Stack
AI-powered campus engine that answers any question using real-time data in a crowdsourced-connected knowledge graph. Piloting at Howard with 10K+ students.
Key Features
Metrics
Tech Stack
LLM-powered knowledge CLI with RAG and structured JSON output. Features local LLM inference, Weaviate vector store, and JSONformer validation for safe programmatic use.
Key Features
Metrics
Tech Stack
Fine-tuned LLaMA 2 7B model using QLoRA + PEFT on 5K+ campus-specific queries. Production inference with BLEU/perplexity evaluation and GPU optimization.
Key Features
Metrics
Tech Stack
Scalable RAG system with Mixture-of-Experts routing. Gated softmax classifier routes queries to domain-specific vector stores and specialized LLMs.
Key Features
Metrics
Tech Stack
Live dashboard for LLM resource monitoring with Prometheus + Grafana. Tracks GPU usage, token throughput, and system health with intelligent alerting.
Key Features
Metrics
Tech Stack
Automated test framework for LLM systems with fuzzing, integration tests, and CI/CD integration. Ensures reliability across inference, RAG, and output handling.
Key Features
Metrics
Tech Stack
Distributed Graph Neural Networks with Byzantine Fault Tolerant consensus. Research implementing topology-aware consensus, adaptive partitioning, and federated GNN training.
Key Features
Metrics
Tech Stack
Offline-first solar-powered AI kit for rural Africa. Runs LLMs and computer vision models locally using Jetson + solar. Built for rural clinics, schools, and farms.
Key Features
Metrics
Tech Stack
"Shazam for video scenes" using CoreML + OpenCLIP + FAISS + on-device ViT. IDs short video clips instantly and deep-links to streaming platforms.
Key Features
Metrics
Tech Stack
Swift app for counter-trend trading using satellite, social, and financial data. TensorFlow + LSTM + real-time inference on iOS.
Key Features
Metrics
Tech Stack
Custom Python trading bot with Random Forests and live execution via Alpaca. 1.8 Sharpe ratio, sub-250ms latency.
Key Features
Metrics
Tech Stack
These projects represent my journey in building AI systems that work in the real world. Each one tackles unique challenges and pushes technological boundaries.
Let's Work Together
Ready to build something revolutionary? Whether it's AI systems, edge computing, or production ML platforms, let's create the future together.