Tushar Garg
Software Engineer building full-stack & AI applications
New to Contra
Tushar is ready for their next project!
Founded and developed FinKizuna ERP, a production-ready multi-tenant ERP SaaS platform for business workflow automation and enterprise management. Built the complete full-stack architecture using React, TypeScript, FastAPI, PostgreSQL, Docker, and AWS. Designed scalable backend APIs, role-based access control (RBAC), workflow automation, and cloud deployment. Integrated AI-powered fraud detection with 94% detection accuracy using explainable AI techniques. Implemented a bilingual (English & Japanese) interface to support international users and financial operations across global and Japanese business environments, contributing across product architecture, development, deployment, and continuous improvement.
1
41
Developed VERITAS, a production-ready environmental intelligence platform analyzing 11,000+ environmental records for real-time indoor air quality monitoring and risk assessment. Built scalable backend services using FastAPI, designed efficient data processing pipelines with Python and Pandas, and implemented automated reporting workflows. Integrated XGBoost with SHAP for explainable AI-powered decision support, contributing across backend development, API design, deployment, and production optimization.
1
30
Explainable HIV Drug Resistance AI : A research-orientated bioinformatics platform that predicts HIV drug resistance from genomic sequences using Python-based ensemble ML (Random Forest + XGBoost) and SHAP explainability, integrated with a Java Spring Boot backend for scalable REST APIs, application services, and system orchestration. Dockerized for reproducible deployment and designed for future clinical and research-scale genomic analysis.
1
12
ZenKensa – 製造検査システム
• FastAPI と TensorFlow Lite を用いた AI 支援型製造検査システムのソフトウェア開発に従事。
• コンピュータビジョンの処理パイプラインを構築し、リアルタイム金属表面欠陥検出のための推論処理を最適化。
• 検査レポートの自動化、バックエンド API、ワークフロー自動化、および意思決定支援機能の開発に貢献。
• 製造現場向けシステムの設計・実装・デプロイメントを支援。
1
15