Engineering the IAC RFQ Automation Tool — replacing Excel-based RFQ workflows with a visual, RBAC-aware React platform.

• Built the frontend architecture in React.js and TypeScript, improving RFQ accuracy by 90% and cutting processing time by 95%.
• Architected document-level RBAC with three access tiers (View / View-Edit / View-Edit-Status Change) across Admin, Employee, and Vendor roles.
• Shipped role-specific layouts, including a fully isolated supplier-facing experience for Vendors.
• Developed RFQ assignment workflows so Sales can allocate employees while restricting document visibility to assigned teammates.
• Visualized complex RFQ and collaboration processes with XYFlow and Excel-like TanStack Table editors.

Founded and shipped a four-language digital news platform — owning product, APIs, caching, AI pipelines, and cloud infrastructure.

• Designed the full stack in Next.js, React, Node.js, PostgreSQL, and Prisma, publishing in English, Hindi, Marathi, and French.
• Implemented Redis caching that cut average API response time from ~400ms to 40–50ms (~90% faster).
• Built an AI translation pipeline so one article auto-publishes across all four languages, plus an LLM cron pipeline generating scheduled content from Excel/CSV title lists.
• Ran background jobs with BullMQ and Redis for translation, scheduled publishing, and content generation.
• Dockerized services, deployed the API on AWS EC2 with media on S3, hosted the frontend on Vercel, and grew organic DAU from ~500 to ~1000 with SSR SEO.