shriprakash tiwari

I’m a Software Engineer with 5+ years of experience building scalable backend systems, real-time applications, and AI-powered products. I focus on distributed systems and dependable infrastructure that keeps consumer experiences fast and resilient.

In my latest role, I engineered an adaptive prompt framework to evolve character behaviors using long-term interaction context and emotional states. I also led database optimization by migrating large Base64 audio payloads from MongoDB to Azure Blob Storage, implementing an ingestion pipeline that compressed uploads with ffmpeg and stored only Blob URLs—reducing database storage by 91% (3.5TB → 310GB).

I architected provider-agnostic payment infrastructure with strictly-typed interfaces and webhook pipelines, enabling payment provider hot-swaps mid-outage without system downtime. I designed a custom WebSocket service with an event-driven named queue system and integrated it across chat and image-generation pipelines to replace inefficient polling with instant, event-driven rendering.

I’ve built search and discovery systems using OpenSearch and improved performance with multi-layer Redis caching for character profiles, chat history, and homepage lookups. I also run data-driven feature rollouts using GrowthBook A/B tests with real-time conversion tracking, and I’ve led teams—like taking ownership of an e-commerce platform with Node.js and MongoDB.