AI & ML Researcher · NASA International Space Apps Champion 2025 · IEEE Member
Building interpretable machine learning systems at the intersection of behavioral modeling, social engineering detection, and socio-technical resilience.
I work at the intersection of AI, behavioral modeling, and socio-technical resilience building systems that are not only accurate but explainable and grounded in human reality.
I'm a BSc Computer Science & Engineering student at Dhaka International University (CGPA 3.76 / 4.00), graduating December 2027. My research focuses on detecting social engineering through interpretable behavioral anomaly modeling, predicting system failures using socio-economic and IoT data, and building continual learning frameworks for dynamic environments.
As President of DIU Computer Programming Club and an active IEEE & IEEE Computer Society Member, I combine research depth with community leadership championed by a NASA Space Apps 2025 victory with Team Polaris.
Fig. 0 — Research Methodology Framework
Accepted, under review, and ongoing research across ML, AI, socio-economics, and computer vision.
Research prototypes and engineering systems spanning ML, CV, and full-stack development.
A comprehensive guide for AI/ML researchers covering research methodology, paper writing, experimentation design, and navigating the academic publication process. Available on ResearchGate.
Fig. 4 — Academic Lifecycle
NASA NEO data platform with real-time 3D visualization. Champion, NASA International Space Apps Challenge 2025 - Barisal Division & Global Honorable Mention with Team Polaris.
Fig. 3 — MeteorShield Pipeline
Distributed, scalable platform integrating LLM agents and resilient proxy routing for automated content generation.
Fig. 5 — Distributed LLM Pipeline
XGBoost model trained on IoT time-series data to predict food spoilage (F1: 0.89). Dashboard visualizes waste reduction impact.
Fig. 6 — Spoilage Prediction Model
TypeScript full-stack engine analyzing developer workflows, integrating GitHub services with burnout tracking algorithms.
Fig. 7 — Burnout Telemetry
ML model identifying high-risk urban segments using traffic patterns and accident history (Precision: 84%).
Fig. 2 — Risk Synthesis
Hierarchical Temporal-Spatial Pattern Fusion model for advanced data analysis and forecasting in dynamic environments.
Fig. 1 — Hierarchical Fusion
AI-powered supplementary educational platform enhancing university coursework with intelligent assistance and interactive tracking.
Fig. 8 — Educational Assistant Loop
Real-time cryptocurrency portfolio management application for tracking live market data and analyzing asset performance.
Fig. 9 — Market API Stream
Technical Stack
Led Team Polaris to victory at the Barisal Division championship. Combined expertise in Machine Learning, full-stack development, data analysis, and UI/UX to engineer a winning planetary defense solution — MeteorShield.
Open for research collaborations, engineering challenges, or just a friendly conversation about AI and its impact on humanity.
meetmehedi1@gmail.com