I architect scalable solutions, design clean APIs, and ship reliable systems. Specialized in Python, Django, and Cloud Infrastructure with a passion for automation and simplicity.
Status: Available
Role: Backend Lead
Designing scalable, high-performance systems using Python & Django. Microservices & Monoliths.
Start Project_Optimizing AWS/Azure infrastructure for cost and performance. Security auditing & automation.
Start Project_Building robust RESTful APIs with fast integration, clear documentation, and secure endpoints.
Start Project_Computational pathology research, data analysis, and implementing predictive machine learning models.
Start Project_Mediusware Ltd.
Developed and maintained web applications, implemented new features, and optimized performance for client projects.
Mediusware Ltd.
Built responsive web interfaces, collaborated with design team, and contributed to agile development processes.
Absolute Ops, a FinOps platform optimizes cloud workloads for performance and cost, with built-in security analysis.
Real estate intelligence platform for U.S. investors to select market and property to invest in best possible way.
Microservices-based backend system for managing user cards, integrating high-speed scraping and Typesense search.
An in-house HRM system for managing Mediusware LTD. employees, payroll, and attendance.
This paper compares Vector AutoRegression (VAR), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) models for forecasting multivariate IMU-based gait data to monitor Parkinson's Disease, finding GRU to be the most accurate model.
This paper proposes a privacy-preserving federated learning framework with a dual-stream ResNetRS50 backbone for multi-scale colorectal cancer histopathological grading, achieving 83.5% accuracy and high recall (87.5%) for the critical Grade III tumors.
This study proposes an interpretable deep learning framework using an EfficientNet-B2 backbone with a prototype-based head and Out-of-Distribution (OOD) detection for multi-class tea leaf disease classification, achieving a balanced accuracy of 97.87% on the Tea LeafBD dataset.
I'm always interested in new opportunities—whether it's a freelance project, large-scale collaboration, or a full-time engineering role.