Technology · 7 clusters · 39 technologies

A production-grade stack for AI-native platforms.

Seven clusters, connected the way the systems are: models and retrieval, the services behind them, the interfaces in front, and the infrastructure underneath. Every name is readable below.

  • AI & LLM

    7

    Models, orchestration and retrieval.

    • LangChain
    • OpenAI / GPT
    • Ollama
    • Whisper
    • RAG Systems
    • Agentic AI
    • Prompt Engineering

  • Vector DB

    4

    Semantic memory for RAG.

    • Pinecone
    • FAISS
    • ChromaDB
    • Weaviate

  • Backend

    6

    APIs, services and business logic.

    • FastAPI
    • Node.js
    • Laravel
    • Flask
    • .NET
    • PHP

  • Frontend

    6

    Product interfaces and websites.

    • React
    • Next.js
    • Vue.js
    • TypeScript
    • Tailwind CSS
    • Bootstrap

  • Cloud & DevOps

    6

    Deployment, pipelines and edge.

    • AWS
    • Docker
    • GitLab CI/CD
    • Lambda
    • S3
    • Cloudflare

  • Database

    5

    Relational, document and cache.

    • PostgreSQL
    • MySQL
    • MongoDB
    • Redis
    • DynamoDB

  • Real-Time & Vision

    5

    Live communication and document vision.

    • WebRTC
    • Socket.io
    • OCR / Vision AI
    • PDF Parsing
    • Multimodal AI

Build tooling

The agentic coding stack

How these are orchestrated
  • Claude CodePrimary agentic pair-programmer
  • CodexOpenAI code generation & review
  • GitHub CopilotInline completion & chat
  • CursorSpec-aware AI IDE
  • Gemini CLILong-context planning agent
  • WindsurfAgentic flow IDE
  • ClineAutonomous coding agent
  • Replit AgentCloud build & deploy agent
  • AiderGit-native AI pair-programmer
  • v0AI UI & component generation

One message is enough

Need a senior architect to choose the stack?

No forms, no booking calendars. Send the idea the way you would to a friend; Kingshuk replies personally.

WhatsApp Kingshuk Email the idea