I'm an Electrical & Electronic Engineer (First Class Honours, University of Khartoum) working at the intersection of AI research and applied software engineering. My work spans two worlds I love equally:
- 🤖 AI & Robotics — LLM-based robotic task planning, Vision-Language Models, and training-free video anomaly detection. Co-author of GridVAD, a zero-shot anomaly detection pipeline achieving state-of-the-art Pixel-AUROC on UCSD Ped2.
- 📱 Full-Stack & Mobile — Building production apps end-to-end with Flutter, React Native, Node.js, and Supabase, from UI to backend services and LLM integrations.
🔭 Currently a Platform Engineer at Enigma AI, alongside a Research Assistant role at the Institute of Space Research & Aerospace and a Teaching Assistant position in Control Engineering. 🌱 Exploring agentic AI systems with LangGraph & LlamaIndex. 💬 Ask me about robotics, computer vision, LLM agents, or shipping cross-platform apps. 📫 Reach me at obadabadee.pro@gmail.com
Languages
AI / ML & Robotics
Frontend & Mobile
Backend & Cloud
Training-free pipeline using Vision-Language Models for zero-shot video anomaly detection. Achieves 77.59 Pixel-AUROC on UCSD Ped2 — best among all compared methods. Combines a propose–ground–propagate architecture (VLM proposals → Grounding DINO → SAM2 mask propagation).
Python PyTorch VLMs SAM2 · 📄 arXiv · 🌐 Project Page
Graduation project (Grade A+). An affordance-guided framework integrating LLMs with low-level robotic control — interprets scene semantics, selects action primitives, and executes on a Franka Emika Panda arm in Webots simulation. Bridges symbolic reasoning and continuous action through a modular perception–reasoning–execution architecture.
Python LLMs Webots Robotics
A fitness-coaching mobile app built with Flutter. Coaches add clients, assign workout tasks, and track their progress, while clients update tasks to log their development — streamlining the whole coaching workflow in one place.
Flutter Dart Mobile
A UNIX command-line interpreter written in C · ALX project. Replicates core shell behaviour — command parsing and execution, PATH resolution, built-ins, environment handling, and error management — built from scratch using low-level system calls.
C Linux Systems Programming
- 📄 Published researcher — Co-author of GridVAD (arXiv:2603.25467)
- 🎓 First Class Honours in Electrical & Electronic Engineering (Control & Instrumentation)
- 💻 IEEE Xtreme 12.0 — 4th place in Sudan, 388th globally
- 🌍 Content Team Coordinator — IndabaX Sudan, the national AI conference
- 🎒 ALX Africa Software Engineering Graduate — 12-month intensive programme in systems programming, C, Python, algorithms, REST APIs, and large-scale software design
- 📚 DeepLearning.AI Machine Learning Specialization (In Progress)
"Bridging research and real-world systems — from robotic arms to mobile apps."