I build intelligent systems that turn complex data into working products.
Portfolio · 2026
My specialties
- AMachine LearningPyTorchDeep LearningTandem NetworksPredictive Modeling
- BComputer VisionOpenCVPose EstimationCNNsFDTD SimulationCLIP
- CNLP & Language ModelsGPT-4oBERTVector SearchTransformers
- DMLOps & ProductionDockerFastAPIGCPRabbitMQClickHouse
- EFull-Stack EngineeringNext.jsTypeScriptNestJSPostgreSQL
- SSentaurus TCAD SimulationsGAAFETDevice PhysicsTCADFinFET
Selected work - 01/05
Recent
projects
IEEE EDTM 2026
Tandem Networks for Metasurface Inverse Design
Turning clicks
into intent.
~Behavioral SignalEvent Pipeline
Open Source
Event-Driven LLM Segmentation
ProvaBook
Production OS for Textile Buying Houses
500 students,
one source of truth.
UIGMISUniversity LMS
IGMIS University
Learning Management System & Grading Engine
Academic output
Research &
publications
IEEE EDTM 2026
ML-Assisted Solar Cell Design via Tandem Neural Networks
IJABC 2024
Nurses' Skill Assessment via Computer Vision
The principles behind my practice
Rigorous.Scalable.Purposeful.
Good ML engineering is more than a high score on a validation set. It means models that survive real data, pipelines that don't quietly fail at 3 a.m., and systems that solve problems people actually have.
I work at the intersection of research rigor and production pragmatism, trained on EEE fundamentals and sharpened by peer review and production users.
Start a conversationPythonPyTorchComputer VisionLLMsNLPMLOpsDockerRabbitMQClickHouseNext.jsNestJSDjangoTypeScriptPostgreSQLFDTDPythonPyTorchComputer VisionLLMsNLPMLOpsDockerRabbitMQClickHouseNext.jsNestJSDjangoTypeScriptPostgreSQLFDTD
Continuing education
Certifications
& courses
Academic grounding
Education
University of Dhaka