Building backend infrastructure for AI-powered and distributed systems.

I'm Vansh Choudhary, a Software Engineer with 1 year of experience building production backend services, distributed systems, AI-powered developer tools, and real-time communication platforms.

I specialize in designing backend architectures that automate complex workflows, coordinate distributed services, and provide reliable infrastructure for modern applications.

Focus: Backend Systems
Domain: AI Infrastructure
Experience: 1 Year

What I build

Backend platforms built for distributed systems, AI infrastructure, and production workloads.

My work spans backend APIs, distributed services, deployment platforms, AI agents, observability systems, and cloud infrastructure.

I solve engineering problems where architecture matters — service communication, asynchronous processing, execution isolation, deployment automation, and system reliability.

Experience 1 Year

Primary Stack Node.js · TypeScript · Redis · Kafka · RabbitMQ · MongoDB · PostgreSQL

Focus Backend Infrastructure • Distributed Systems • AI Platforms

Selected Work

Systems built to automate engineering workflows and production infrastructure.

TUI Terminal Code

A terminal-native AI coding agent that combines autonomous tool execution, browser automation, file operations, and multi-model reasoning into a production-style developer workflow.

Autonomous tool execution
Playwright browser automation
Multi-provider LLM support
Long-running conversation memory
Shell execution
File editing
Sub-agent orchestration

Category: Developer Tooling
Architecture: AI Agent Loop
Language: TypeScript

Repository

DeployForge

Transforms GitHub repositories into production deployments through distributed workers, Docker build pipelines, and AWS infrastructure.

Docker build pipeline
Distributed deployment workers
EC2 backend deployment
S3 frontend deployment
Redis Pub/Sub
RabbitMQ orchestration

Architecture: Distributed Workers
Cloud: AWS
Language: TypeScript

Live Repository

DebugPilot

An AI-powered observability platform that correlates production logs, service metrics, deployments, and repository code to generate asynchronous root-cause analysis.

Event-driven incident processing
Asynchronous BullMQ RCA workers
Repository-aware RAG pipeline
MongoDB Vector Search & embeddings
Incident fingerprinting & memory
Service metrics & deployment correlation

Architecture: Event-Driven
Domain: AI Observability
Language: TypeScript

Live Repository

Pi Sandbox Kubernetes Runtime

A secure execution runtime that isolates every AI tool invocation inside disposable Kubernetes sandbox pods for safe, session-aware code execution.

Kubernetes sandbox execution
Warm pod pool
Disposable execution environments
Filesystem isolation
Session-aware runtime
Secure tool execution

Infrastructure: Kubernetes
Domain: Secure Execution
Language: TypeScript

Live Repository

RAG Optimization

A modular Retrieval-Augmented Generation pipeline implementing advanced chunking, embeddings, and retrieval optimization for higher-quality LLM responses.

Semantic & adaptive chunking
Query transformation
Embedding generation
Vector search
Retrieval evaluation
Modular RAG architecture

Domain: AI Infrastructure
Category: Retrieval Pipeline
Language: TypeScript

Repository

Technical Stack

Technologies used across backend platforms, cloud infrastructure, and AI systems.

Languages

TypeScript, JavaScript, C

Backend

Node.js, Express.js, REST APIs, Background Job Processing

Databases

MongoDB, PostgreSQL, SQL, Redis, MongoDB Vector Search

AI

OpenAI, Gemini, RAG, Embeddings, Tool Calling, AI Agents, MCP

Distributed Systems

Kafka, RabbitMQ, Event-Driven Architecture

Infrastructure

Docker, Kubernetes, AWS, GitHub Actions, Nginx

Realtime

WebRTC, Mediasoup, WebSockets

Find me online

GitHub vansh-choudhary01 LinkedIn vansh-choudhary X @VanshChoudhaary