I'm an AI/ML Engineer and Data Science student at Southeastern Louisiana University, expecting to graduate in May 2026. I build and deploy intelligent systems, specializing in computer vision, edge AI optimization (like quantization for NVIDIA Jetsons), and production-ready RAG pipelines.
-
Student Engineer @ Southeastern Northshore STEM Center (Aug 2025 – Present)
- Lead technical enablement workshops, translating complex hardware architecture and AI concepts for non-technical audiences to drive customer engagement.
- Engineered computer vision pipelines for NVIDIA Jetson-powered robots, applying quantization to optimize custom machine learning workflows for real-time edge processing.
- Designed and deployed autonomous navigation systems using PyTorch models, benchmarking object detection algorithms to showcase hardware capabilities to stakeholders.
-
AI Intern @ Dialysis Care Center (May 2025 – Aug 2025)
- Engineered production-ready RAG pipelines using PyTorch to streamline patient interaction workflows, ensuring strict adherence to healthcare data governance and compliance standards.
- Architected a semantic vector search engine (FAISS) with optimization techniques to modernize legacy knowledge retrieval, significantly reducing query resolution time for medical staff.
- Collaborated with cross-functional teams to integrate LLM-based autonomous agents, implementing guardrails to eliminate hallucinations in high-stakes clinical environments.
-
Research Assistant @ Southeastern Louisiana University (Aug 2024 – Dec 2024)
- Implemented end-to-end data acquisition protocols and preprocessing pipelines for high-fidelity EEG signals, ensuring reproducibility of scientific results.
- Trained and evaluated classical ML classifiers (SVM, KNN, Decision Trees) to validate novel biometric identifiers, applying advanced data augmentation to handle sparse biological datasets.
| Project | Description | Stack |
|---|---|---|
| Inferno Fire & Smoke | Developed a commercial-grade fire and smoke prevention system by finetuning RF-DETR and YOLO11n architectures on a robust dataset of 160k+ real and augmented images. Applied post-training quantization to achieve ultra-low latency inference natively on edge devices. | RF-DETR, YOLO11, Edge AI |
| PragBase | Architected a universal RAG assistant using Flask and HTMX that ingests agnostic data sources, converting them into a vector-searchable knowledge base. Developed a lightweight widget to embed the AI agent into external websites. | Flask, HTMX, RAG |
| Paokinator | Engineered a probabilistic guessing game utilizing a FastAPI backend and PostgreSQL to query over 260k+ datapoints. Implemented a self-correcting fuzzy logic algorithm that updates decision weights based on real-time feedback. | FastAPI, PostgreSQL |
ML / AI
Web, Backend & Infrastructure
Languages
Open to internships and collaborative projects in AI/ML. olisemeka.dev · Download Resume
