19-year-old software engineer focused on backend systems, AI agents, and real-time applications.
I primarily work with TypeScript, Node.js, Next.js, Redis, PostgreSQL, Docker, and WebSockets — building AI workflows, realtime systems, and production-style architectures beyond simple CRUD apps. Currently exploring scalable AI agent infrastructure, MCP integrations, queues/workers.
there are something i believe
work harder for your dreams
whatever happens, happens
dont worry about things you can't control
view moreGitHub Activity
Stack
projects
currently backend service are suspended due to free tier
guarded ai agent
Production-style AI agent platform with MCP server support, dynamic tool discovery, policy guardrails, realtime approvals, Dockerized infrastructure, and AI tool orchestration. The system allows LLMs to securely interact with tools through a governed execution pipeline with audit logging and human approval workflows.
I got an assignment https://docs.google.com/document/d/1bbsYwJNfjm-o4IoNkQQ3wLGXJyWgE_Sd2ywLRH83zqQ/edit?tab=t.0
Accerra
Accerra a jee neet web application that gave The unfair advantage for serious JEE aspirants.
Accerra is a student intelligence system designed to understand how a student understands concepts — not just whether answers are right or wrong. Through its Activity, Intelligence, and Teaching layers, Accerra analyzes PYQ performance levels, note-taking patterns, learning behavior, and study consistency to build a structured long-term student brain model. Instead of relying only on generic AI conversations, Accerra combines structured cognition tracking, decision engines, and RAG-based contextual memory pipelines to give the AI deeper awareness of how each student thinks, learns, and improves over time. This enables a deeply personalized, NCERT-grounded JEE/NEET preparation experience that adapts to the student rather than forcing every student into the same learning approach.