ContractJan, 2025 - Present

Independent Contractor

Full Stack Engineer

📍 India - Remote

Overview

Working as an independent contractor, I've been building enterprise-grade SaaS platforms and AI-powered solutions for clients across fintech and media industries. My focus has been on delivering scalable, production-ready applications with modern tech stacks.

Technologies Used

Next.jsNext.js
ReactReact
TypeScriptTypeScript
FastAPIFastAPI
PythonPython
PostgreSQLPostgreSQL
AWSAWS
DockerDocker
RedisRedis
TanStack QueryTanStack Query
ZustandZustand

Projects

Risk Prediction Platform

Completed

A multi-tenant SaaS platform for financial institutions to assess company default probability using ML-powered models. The platform enables real-time risk assessment and bulk portfolio analysis.

  • •Built a multi-tenant architecture supporting multiple financial institutions with complete data isolation
  • •Designed and implemented role-based access control (RBAC) with JWT authentication across tenants, organizations, and users
  • •Developed ML inference pipelines using Logistic Regression and LightGBM ensembles for annual and quarterly predictions
  • •Created async APIs for real-time predictions and bulk portfolio risk analysis with CSV/Excel processing
  • •Built high-performance backend with FastAPI, PostgreSQL, Redis, and Celery for background ML jobs
  • •Implemented real-time analytics dashboard with Next.js, TanStack Query, and Zustand for interactive insights
  • •Architected cloud infrastructure on AWS (ECS Fargate, RDS, Redis, ALB) with CI/CD pipelines using GitHub Actions

AI Data Analytics Platform

In Progress

An AI data science chatbot platform that combines conversational analytics with notebook-style Python execution for real-time insights.

  • •Built an intelligent chat interface in Next.js 16 using AI SDK 5.0 and Anthropic Claude to translate natural language questions into executable analysis workflows.
  • •Integrated E2B Code Interpreter to spin up isolated Python notebook environments with pre-installed data-science libraries (pandas, numpy, matplotlib, seaborn, plotly, scikit-learn).
  • •Implemented file upload and persistence using Vercel Blob and PostgreSQL (Drizzle ORM), making uploaded CSV/XLS files available inside notebook sessions.