The Cipher Stack

I’m Shahanur Islam Shagor, a Full-Stack Web Developer and AI / Autonomous Systems Engineer focused on building dependable software and intelligent systems that connect the web, AI, edge computing, and the physical world.

My work spans full-stack product engineering, machine learning, computer vision, autonomous systems, distributed architectures, and applied cybersecurity. I enjoy taking complex problems from idea to implementation—designing the architecture, building the product, integrating intelligent components, validating performance, strengthening security, and shipping systems that are practical to use and maintain.

Across my projects, I have worked on scalable web platforms, local-first AI applications, edge-AI systems, UAV navigation and sensor fusion, privacy-preserving machine learning, secure connected systems, and research-driven engineering prototypes. What matters most to me is not adding technology for its own sake, but choosing the right engineering approach for the problem and making the result reliable in real-world conditions.

I approach engineering with a strong product and systems mindset: clean architecture, secure-by-design decisions, measurable performance, maintainable code, thoughtful UI/UX, automation, and reliable deployment. I’m especially interested in work where software must do more than function—it needs to be useful, resilient, scalable, and technically sound.

• Based in Voronezh, Russia · UTC+3
• Portfolio: https://smshagor.com
• GitHub: @smshagor-dev
• Email: smshagor.dev@gmail.com
• Core Focus: Full-Stack Engineering · AI/ML · Autonomous Systems · Edge AI · Applied Security

Contribution data is generated automatically by the profile artwork workflow and refreshed daily.

Automatically synchronized from my public ORCID record every 6 hours. Last sync: 05 Sep 2026, 20:27 UTC.

No works currently classified by ORCID as published/formal outputs.

1. Adaptive Edge Intelligence for Low-Latency Assistive Vision under Dynamic Network and Computational Constraints — 2026-09-03 · Preprint
2. Fabrication-constrained tandem inverse design of photonic crystal fiber SPR sensors with uncertainty-aware Pareto screening — 2026-09-03 · Preprint
3. Risk-Aware Adaptive Video Processing with Freshness Control for Low-Latency Assistive Vision — 2026-09-03 · Preprint
4. A Secure Offline Edge Architecture for Wearable Vision Assistance: Raspberry Pi-ESP32-CAM Integration and Reproducible Software Validation — 2026-09-02 · Preprint
5. WVAB: A Low-Latency Wireless Assistive Vision Framework for Risk-Aware Navigation and Multilingual Audio Guidance — 2026-08-31 · Preprint
6. Toward Blockchain-Assisted Zero-Trust Secure Communication for Decentralized UAV Swarms in GPS-Denied Environments: A Mathematical Security and Scalability Framework — 2026-08-27 · Preprint
7. Cyber-Resilience and Trust-Aware Command Authorization for Autonomous UAV Swarms in GPS-Denied Environments — 2026-08-26 · Preprint
8. Decentralized Coordination and Acoustic Source Localization in Autonomous Drone Swarms — 2026-08-26 · Preprint
9. OmniGuard V2X: A Hybrid-Security Prototype Framework with Assumption-Aware Validation for Smart Vehicle Systems — 2026-08-26 · Preprint
10. Policy-Gated Firmware Trust and Anti-Rollback Resilience for Autonomous UAV Edge Platforms: A Formal Security and Software Validation — 2026-08-26 · Preprint
11. Sequence-, Epoch-, and Integrity-Aware Detection of Replay and Peer-Spoofing Attacks in Decentralized UAV Mesh Networks: A Software-Validated Security Framework — 2026-08-26 · Preprint
12. Toward Secure Decentralized Coordination and Acoustic Source Localization in Autonomous Drone Swarms — 2026-07-31 · Preprint

A curated selection of my strongest work across AI systems, autonomous platforms, cybersecurity, privacy-preserving ML, edge intelligence, real-time software, and production-grade full-stack engineering.

OpenMindAI A local-first desktop AI workstation for private on-device inference, model and runtime management, agents, workspaces, portable storage, media workflows, and offline-capable intelligent assistance.	ZeroTrust-FL-Sim A zero-trust federated-learning security testbed combining Byzantine-robust aggregation, adversarial clients, differential privacy, CKKS, post-quantum transport, SIMD/CUDA acceleration, chaos engineering, and full observability.
UAV GPS-Denied Navigation Autonomous UAV navigation for GPS-denied and dynamic environments using resilient localization, VIO, ESKF-based sensor fusion, UWB/TDOA, LiDAR-aware perception, fault injection, and deterministic validation workflows.	AetherMotion A C++-first real-time gesture intelligence and spatial AR engine with native vision pipelines, gesture classification, smoothing, spatial interaction, replay, telemetry, and low-latency runtime execution.
Next-Gen V2X Security Applied vehicle-security research combining decentralized trust, modern cryptography, authenticated communication, post-quantum security concepts, and resilient architecture for connected mobility systems.	Secure Autonomous Drone Swarms Multi-agent autonomy research spanning decentralized coordination, acoustic localization, secure communication, distributed decision-making, swarm resilience, and cooperative UAV behavior.
WVAB — Wireless Vision Aid A low-cost edge-computer-vision assistive platform for real-time scene understanding, object detection, risk-aware navigation, wireless processing, and multilingual audio guidance for visually impaired users.	CyberPhotonics-SPR Research-oriented engineering at the intersection of photonic sensing, cyber-physical security, resilient system design, experimental validation, and software-backed secure sensing workflows.
Federated Learning on Non-IID Data + Differential Privacy Privacy-preserving federated-learning experimentation for heterogeneous non-IID environments, combining distributed model training, differential privacy, evaluation, and practical ML analysis.	B2B Global Trade A production-oriented multi-tenant export/import marketplace with RFQs, quotations, supplier verification, live chat, advanced search, payments, 2FA, RBAC, analytics, queues, and Docker-based deployment architecture.
SyncChat A modern real-time communication platform focused on synchronized messaging, responsive cross-device UX, production-oriented communication flows, and scalable full-stack architecture.	CodeCraft An advanced browser-based development environment with Monaco Editor, multi-file projects, a virtual file system, live preview, console tooling, project templates, import/export, and responsive developer workflows.

A work-aligned stack built from the technologies I use across production web platforms, local-first AI, mobile and desktop applications, autonomous systems, edge computing, applied security, distributed systems, and research engineering.

TypeScript · JavaScript · Python · C++20 · Go · PHP · Rust · Dart · Kotlin · Java · Bash · PowerShell

HTML5 · CSS3 · React · Next.js · Tailwind CSS · Bootstrap · Material UI · Sass / SCSS · Vite · Three.js · Figma

Node.js · Express.js · Laravel · Strapi · PostgreSQL · MySQL · MongoDB · Redis · SQLite · Prisma ORM · Firebase

PyTorch · TensorFlow · Scikit-learn · OpenCV · Flutter · Android · Tauri · Raspberry Pi · Arduino · Qt · CMake

Docker · Nginx · Linux · Git · GitHub · GitHub Actions · AWS · Google Cloud · Kubernetes

_{Pull Shark ×3 · YOLO · Starstruck · Quickdraw}

_{Scan to open my portfolio}

Build. Research. Secure. Automate.

_{Designing software and intelligent systems for the web, the edge, and autonomous environments.}