Hi there, I'm Shahriar Alom Masud 
- Role: Embedded Systems & Edge AI Engineer
- Education: B.Sc. in IoT & Robotics Engineering — University of Frontier Technology, Bangladesh
- Currently: Actively building systems and continuously developing myself
- Open Source: Author of SmartProv (v2.1.3) on official Arduino Library Manager
- Core Values: Clarity, ownership, fault-tolerance, and continuous improvement
- Contact: shahriar0002@std.uftb.ac.bd • Dhaka, Bangladesh
A Perspective on Failure & Growth
Bad News:
No matter how hard you try to protect yourself, one day, you will fail.
Good News:
When you fail, each passing day, it opens you up to new possibilities, often helping you emerge as the best version of you.
|
|
|
|
|
|
|
Official Arduino Library Manager Package An asynchronous, event-driven captive-portal Wi-Fi provisioning framework for ESP32 and ESP8266 microcontrollers. Engineered non-volatile NVS flash persistence with automated fallback routines, eliminating manual credential flashing across 100% of device deployments.
🔗 View Repository • 📖 OTA Companion |
👁️ ProctorAIFull-Stack AI Classroom & Exam Proctoring Platform Real-time computer vision platform tracking eye gaze (EAR), mouth ratio (MAR), phone presence, and 3D head pose via solvePnP at 30–60 FPS. Validated across 35+ students and 3,500+ telemetry records with FastAPI and Supabase PostgreSQL backend.
|
|
Explainable Edge AI Respiratory Risk Monitoring System Investigational edge healthcare platform fusing ESP32 environmental sensing (PMS5003 laser particulate, MQ-135, DHT22) running FreeRTOS with outdoor meteorology. Uses dynamic TreeSHAP local waterfall attributions, interactive Leaflet dispersion mapping, and Groq LLaMA 3.3 AI copilot.
🔗 View Repository • 🌐 Live App |
Risk-Aware Smart Home Energy Management System Time-series forecasting pipeline predicting solar yield and household load via quantile regression and empirical uncertainty margins (kσ) to calculate Safe Solar Surplus. Deployed with dual-layer TreeSHAP explanations and ESP32 discrete RMS power telemetry.
|
|
Explainable Brain Tumor Diagnostic Platform Deep Convolutional Neural Network classifying brain MRI scans into 4 categories with >90% accuracy. Optimized with ONNX Runtime to slash inference latency by ~4×, featuring Grad-CAM clinical heatmaps and automated PDF diagnostic reporting.
|
Autonomous IoT Catamaran & Aquaculture AI System Autonomous smart boat engineered with dual-controller architecture (ESP32 + Arduino Mega 2560), long-range LoRa (SX1278) telemetry, automated servo feeding, InfluxDB Cloud time-series storage, and TreeSHAP-explained fish species suitability prediction.
|
|
Fail-Safe Remote Firmware Delivery & Rollback Architecture Automated over-the-air (OTA) update pipeline using GitHub Releases and HTTPS metadata. Built with dual-partition boot validation, NVS failure tracking, and automatic rollback routines tested across 15+ physical ESP32 nodes with zero bricking.
|
🤖 LinkedIn Automation Workflow ⭐ 11Autonomous AI Content Publishing Pipeline Event-driven n8n automation pipeline that orchestrates AI models for topic discovery, caption drafting, image synthesis, and scheduled direct-to-LinkedIn publishing, cutting manual creation overhead by ~80%.
|
"Passionate about turning hardware signals into intelligent systems and clean code into real-world impact."