Specializing in end-to-end architectures, I design and build high-performance backend systems, secure APIs, and intelligent applications. I bridge the gap between complex data logic and user-centric design to deliver scalable, secure, and modern digital solutions.
An AI-powered insurance claims platform spanning mobile, web, and ML infrastructure, enabling automated damage assessment and multi-role verification workflows.

MOBILE APPLICATION (Flutter): • Real-time AI damage detection with confidence scoring (75%+ accuracy). • Generate customizable, multi-format PDF reports for insurance submissions. • Offline-first architecture with automatic cloud synchronization. WEB PORTAL (Node.js/Express): • Multi-role verification system (Admin, Car Co, Insurance Co) with Supabase integration. • Real-time document upload, review, and approval workflows. • Comprehensive audit trail logging for security compliance. ML/CV API (Flask): • Detectron2 models for vehicle part and damage segmentation. • Concurrent inference using ONNX Runtime with thread-safe serving.
Avoiding monolithic bottlenecks. I focus on orchestrating decoupled architectures—building optimized, independent layers across web dashboards, mobile clients, and central backend API gateways to ensure seamless state synchronization and long-term maintainability.
Going beyond basic database storage. My development process prioritizes integrating advanced utility—whether that means implementing complex backend data cross-matching algorithms or deploying cloud-hosted computer vision pipelines to solve high-impact, real-world problems.
Strict adherence to open-source tool optimization, performance tuning, and database hardening. I build with an explicit focus on secure transactions, fast response times, and clean, responsive UI configurations that render perfectly on any device.
I am always open to discussing new software architectures, open-source projects, or engineering opportunities.