Log in to ArmanOS
- Use the shell to find your way.
- Explore my journey.
- Talk to Morpheus. He’ll guide you through questions about me and my work.
- Race the evolutionary AI drivers I trained at Istanbul Park.
Nokia Bell Labs · Applied AI Researcher · 2026–present
I build semantic edge-cloud digital twins for urban mobility, connecting computer vision with real-time scene reconstruction.
NEXTCONLAB / Smart Connected Vehicles Innovation Centre · Applied AI Researcher · 2026
I worked on autonomous drones and UGVs: localization, multi-agent path finding, and RF geolocation.
Huawei Canada · Applied AI Researcher · 2025
I developed PANAMA, a network-aware multi-agent reinforcement learning framework, and worked on hallucination detection and mitigation for multi-agent LLM systems.
| Project | Focus |
|---|---|
| Ottawa Sionna RT Cellular Digital Twin | Tracing cellular signals through Ottawa with NVIDIA Sionna RT. |
| Skate 3 Rust — Apple Silicon | Bringing the open-source Skate 3 Rust Engine to native ARM64 macOS. |
| GSANet-DCNN | Separating objects in video with guided slot attention. |
| Developer Performance Tracker | GitHub analytics and pull-request review with LLMs. |
| One-Pixel Adversarial Attack | Seeing how much a single pixel can change an image classifier’s answer. |
| LLM Bias Reduction | Exploring fairness and bias reduction in LLMs and classifiers. |
- Semantic Edge-Cloud Communication for Real-Time Urban Traffic Digital Twins — Nokia Bell Labs · IEEE GLOBECOM 2026, accepted.
- PANAMA: A Network-Aware MARL Framework for Multi-Agent Path Finding in Digital Twin Ecosystems — Huawei Canada · IEEE GLOBECOM 2025, first author.
- Methods and Apparatus For Detecting and Mitigating Hallucinations in Autonomous Multi-Agent Systems — Huawei Canada · U.S. patent pending.
Nine evolutionary AI rivals. Three laps. Same car, same rules. See you at Istanbul Park.
Race on ArmanOS · Under the hood
Find me on LinkedIn or Google Scholar, or send me a note.