Loading profile…
David Itina
Open to opportunitiesFounder
I build production-ready AI systems for fraud detection, document automation, and real-world ML workflows.
NigeriaFounder at ITINA AI
About
I've built production-oriented AI systems spanning document automation and offline-payment fraud detection. My Dokusure platform automates customs document extraction for Nigerian clearing agents, while my QPay fraud system reached 99.1% validation accuracy and was deployed as a FastAPI scoring service.
At KodeHauz, I built and deployed a real-time sign-language recognition system using CNNs, MobileNetV2, MediaPipe, and speech output across seven gesture classes. I also moved machine-learning models from notebooks into production-facing Python and FastAPI APIs, and led an AI, Python, web development, and Arduino robotics curriculum for more than 150 students.
I'm building ITINA AI and exploring practical RAG and agentic workflow applications across healthcare, manufacturing, finance, and logistics.
Experience
Founder
Building an AI systems company, evaluating problem spaces across healthcare, manufacturing, finance, and logistics using RAG and agentic workflow techniques.
RAGAI Agentic WorkflowLlamaIndexPrompt EngineeringVector DatabasesFaissPineconeWeaviate
AI/ML Engineer Research Intern
Built and deployed a real-time sign-language recognition system using CNN and MobileNetV2 with MediaPipe hand-tracking and gTTS speech output. Trained classical ML models with PCA-based dimensionality reduction and shipped FastAPI backend services for production-facing APIs.
PythonfastAPIPyTorchTensorFlowKeras
Education
Covenant University
Pursuing a Bachelor of Science in Industrial Mathematics with coursework in linear algebra, probability and statistics, multivariable calculus, and numerical optimization applied to machine learning model design and evaluation.