I’m a Software Engineer and Master’s student at the University of New Haven (4.0 GPA), specializing in Applied AI, Distributed Systems, and Data Engineering. Currently, I serve as a Graduate Research Assistant at the SAIL Lab evaluating AI document-understanding systems, and recently worked as an AI Software Engineer Intern — Lead Backend & Data Pipeline Engineer at Eastern Metal Works building business-critical ETL pipelines and natural-language SQL assistants. Previously at KNUST, I developed high-performance APIs and data platforms serving over 20,000 active users. I am passionate about building scalable, intelligent systems—ranging from backend microservices to offline AI agents and mobile applications. I am actively seeking roles where I can deliver impactful, AI-driven products at scale.
Building Intelligent Systems.
Software & Applied AI Engineer
About Me
-
Microservices
-
AWS / Cloud Native
-
Computer Vision
-
Distributed Systems
-
DevOps & CI/CD
-
AI Agents (PyTorch)
-
System Design
My Projects
Hot & Kool Essentials
Production E-Commerce Platform
Built and deployed a customer-facing fragrance e-commerce platform supporting product discovery, carts, orders, promotions, shipping, authentication, and payment processing.
Implemented atomic payment transactions, idempotent Paystack webhooks, HMAC-SHA512 signature validation, role-based access control, and CSRF protection.
Python, Django, PostgreSQL, Tailwind CSS, Paystack
→Susu Books
Voice-First AI Ledger
Built a voice-first, entirely offline AI ledger application for informal market traders using Gemma 4 and Ollama for edge-based inference.
Implemented a highly decoupled architecture with a Next.js PWA frontend, FastAPI backend, and zero-shot schema extraction to map messy speech to inventory.
Python, FastAPI, Next.js, SQLite, Gemma 4, Ollama
→SentinelMD - Safety Copilot
Clinical Safety AI
Built offline-capable clinical safety tool leveraging MedGemma open-weight models to identify workflow risks from structured and unstructured inputs.
Python, MedGemma, Machine Learning, Systems Design
→Hot & Kool Essentials
Production E-Commerce Platform
Built and deployed a customer-facing fragrance e-commerce platform supporting product discovery, carts, orders, promotions, shipping, authentication, and payment processing.
Implemented atomic payment transactions, idempotent Paystack webhooks, HMAC-SHA512 signature validation, role-based access control, and CSRF protection.
Python, Django, PostgreSQL, Tailwind CSS, Paystack
→Savantly
MLB Intelligence Mobile App
Built a native iOS baseball intelligence application featuring real-time score dashboards, live game tracking, nine-inning linescores, and player and team comparison workflows.
Developed reusable SwiftUI components and NavigationStack-based flows, integrated live sports data, and validated features through iterative sprint reviews.
Swift, SwiftUI, iOS, REST APIs, Sports Data
→Healing Stones - 3D Recon
Archaeology x AI
Reconstructing fragmented 3D archaeological artifacts using digital scans, geometry, and classical machine learning.
Implementing end-to-end pipeline for mesh loading, preprocessing, and robust registration of cultural heritage objects.
Python, Open3D, Machine Learning, 3D Geometry
→RenAIssance - OCR Pipeline
Historical Document AI
Built a CNN-RNN character recognizer with lexicon-constrained beam search to decode noisy 17th-century printed Spanish texts.
Integrated Gemini LLM for automated OCR post-correction, resolving period-specific spelling and ligature errors.
Python, PyTorch, TrOCR, Gemini API, Computer Vision
→Cognitive Flexibility Under Load
AGI Benchmark (Kaggle)
Built an LLM evaluation benchmark measuring executive function (inhibitory control, cognitive flexibility) through parametric rule-switching tasks.
Python, Kaggle Benchmarks SDK, DeepMind AGI Hackathon
→Avera - Safety Analytics
Geospatial Intelligence
Built production web app delivering US neighborhood safety snapshots by combining crime data, environmental alerts, and geospatial context.
Django REST, Next.js, PostGIS, Mapbox, Docker, Redis
→Edge Persona Simulator
Cloudflare AI
Real-time AI agent system simulating edge infrastructure (WAF, Cache, Bot Mgmt) responses to incident scenarios.
TypeScript, Cloudflare Workers, Workers AI
→Tower Defense AI Agent
Reinforcement Learning
Implemented Q-Learning and PPO agents, achieving a 78% win rate versus 62% A* baseline through RL optimization.
Python, PyTorch (PPO), Pygame
→Ticketing & Workflow System
Internal Platform
Built an internal service-request platform supporting ticket submission, routing, status tracking, approvals, reporting, auditing, and operational workflow management.
Integrated retrieval-assisted response suggestions to help users resolve recurring issues while preserving ticket history and reusable institutional knowledge.
Python, Django, PostgreSQL, REST APIs, RAG
→Scalable ML Microservice
MLOps Architecture
Operationalized a Machine Learning API containerized for high-availability deployment via Kubernetes cluster.
Kubernetes, Docker, Flask, CircleCI, AWS
→CertEasy (SaaS Platform)
Completed
Architected a certification prep SaaS platform. Integrated Stripe for subscription management and Google OAuth for secure identity management.
Django, PostgreSQL, Redis, Stripe API, Docker, Tailwind
→Samaritan: Real-Time Chat
Completed
Built a real-time messaging platform using Firebase Firestore for instant data syncing.
React, Firebase NoSQL, UUID, State Management
→
