Software Engineer building full-stack, backend & cloud systems with TypeScript, Go, Node.js, React and AWS.

Tanveer Singh  • Peerlist

BrainRot Lang is an interpreted programming language I built from first principles in Go to understand what happens between source code and program execution.The implementation follows the complete pipeline:Source → Lexer → Parser → AST → Tree-walking InterpreterI wrote a DFA-based lexer, recursive-descent parser, typed AST, evaluator, and scope-chain environment without using compiler frameworks. The language supports variables, loops, recursion, arrays, closures, lexical scoping, first-class functions, and compound operators. Its behaviour is validated through 19 automated test suites.I also built and published a VS Code extension with syntax highlighting, snippets, hover documentation, and a custom file type. GitHub Actions automatically creates binaries for Windows, macOS, and Linux, while a Next.js website provides documentation and installation guides.- [GitHub](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/Tanveer-rajpurohit/brainrot-lang">https://github.com/Tanveer-rajpurohit/brainrot-lang</a>)- [Documentation](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://brainrot-lang-web.vercel.app/">https://brainrot-lang-web.vercel.app/</a>)- [Case Study](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://tanveer-singh.vercel.app/work/brainrot-lang">https://tanveer-singh.vercel.app/work/brainrot-lang</a>)

BrainRot Lang

CoNestify is a real-time collaboration platform I built to understand the systems behind modern communication products rather than only using existing real-time services.The application includes workspaces, public and private channels, instant messaging, shared documents, collaborative lists, a virtual canvas, invitation links, and group audio/video calls.For group calls, I implemented a custom Selective Forwarding Unit using WebRTC and MediaSoup instead of relying on a peer-to-peer mesh. The SFU was load-tested with more than 500 concurrent streams at sub-200 ms latency.Redis supports caching and real-time state synchronization, PostgreSQL stores persistent workspace and messaging data, and the services were containerized with Docker and deployed on AWS.- [Case Study](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://tanveer-singh.vercel.app/work/conestify">https://tanveer-singh.vercel.app/work/conestify</a>)- [GitHub](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/Tanveer-rajpurohit/CoNestify">https://github.com/Tanveer-rajpurohit/CoNestify</a>)

CoNestify

TaxCopilot is an AI-assisted platform for analysing Indian tax notices and preparing legal drafts. I worked across the frontend, backend, API gateway, data layer, and AI workflow.The system is split into four services: a Next.js client, Node.js and Prisma backend with JWT authentication, a rate-limited API gateway, and a separate AI service.AWS Textract extracts text from scanned documents, while PostgreSQL and pgvector retrieve relevant passages from Indian tax statutes. After generation, a citation-validation layer checks every referenced legal section against the retrieved material. An unverifiable citation triggers a warning and an automatic retry instead of silently returning an unsupported answer.The project qualified for the prototype round of the AWS AI for Bharat Hackathon.- [Case Study](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://tanveer-singh.vercel.app/work/taxcopilot">https://tanveer-singh.vercel.app/work/taxcopilot</a>)- [GitHub](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/orgs/TaxCopilot/repositories">https://github.com/orgs/TaxCopilot/repositories</a>)

TaxCopilot

TaskFlow is an asynchronous task-processing platform I built to explore how background-job systems behave under load and how they can be deployed and operated reliably.The API accepts tasks and returns HTTP 202 in under 10 ms, while Redis BLPOP workers process jobs independently in the background. During load testing, the system handled a spike of 1,000 tasks with zero dropped jobs.I deployed the services on Kubernetes and configured HPA to scale workers from 2 to 50 replicas based on demand. Argo CD and Kustomize provide GitOps-based delivery, while GitHub Actions builds multi-stage Docker images. Prometheus and Grafana expose API, queue, worker, and cluster metrics.Key engineering areas: asynchronous processing, queue workers, autoscaling, containerization, GitOps, CI/CD, and observability.- [Application Repository](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/Tanveer-rajpurohit/taskflow">https://github.com/Tanveer-rajpurohit/taskflow</a>)- [Infrastructure Repository](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/Tanveer-rajpurohit/taskflow-infra">https://github.com/Tanveer-rajpurohit/taskflow-infra</a>)- [Case Study](<a target="_blank" rel="noopener noreferrer nofollow ugc" class="break-all link-break text-gray-500 hover:underline" href="https://tanveer-singh.vercel.app/work/taskflow">https://tanveer-singh.vercel.app/work/taskflow</a>)