FIG_001
National Flag of India

Harshit Nakrani

Fullstack AI Engineer.

Overview

Fullstack AI Engineer

he/him

Social Links

GitHub Contributions

Hello

I’m Harshit Nakrani — a full-stack and backend developer focused on building high-performance, scalable backend systems, clean API platforms, and robust product architectures.

Passionate about exploring modern tech, system design, and AI agent frameworks to build developer-centric software.

Currently building Kyronix — the agentic execution layer for modern software, and NOMAD-1 — an autonomous terminal agent.

Blog(6)

Labs(4)

Stack

Backend / Run-time
Backend / Framework
Database
Database / Cache
DevOps / Infrastructure
Frontend

Experience

Kyronix

Location
Surat, India
Location type
(Remote)
Employment status
Current
  • Designing and building the core agentic runtime execution layer.
  • Implementing context compilation, memory management, planning algorithms, and tool execution engines.
  • Designing developer-focused SDKs and clean integration APIs for agent orchestration.
  • TypeScript
  • Node.js
  • Agentic AI
  • API Design
  • SDK Design
  • System Architecture

Freelance

Location
Surat, India
Location type
(Remote)
  • Built NOMAD-1 - an autonomous terminal agent that executes plans using validated JSON-based operations.
  • Developed Shruti AI - a distributed RAG learning companion transforming document streams into knowledge vectors.
  • Developed Playnet - a full-stack streaming platform featuring a custom job queue and Dockerized FFmpeg workers.
  • Built Prepify - an AI-powered voice interview preparation platform simulating interview questions dynamically.
  • Node.js
  • NestJS
  • Express
  • Next.js
  • React Native
  • PostgreSQL
  • Redis
  • Docker
  • FFmpeg
  • MongoDB

Education

  • TypeScript
  • Node.js
  • NestJS
  • Docker
  • Redis
  • PostgreSQL
  • MongoDB
  • Algorithms
  • Systems Design
  • Software Engineering

Projects(6)

  • Kyronix is an AI Agent Infrastructure Platform providing the complete agentic execution layer between applications and LLMs.

    • Manages planning, memory, context compilation, tool execution, retrieval, workflows, and observability.
    • Bypasses rebuilding agent orchestration loops for every AI app.
    • TypeScript
    • Node.js
    • Agentic AI
    • API Platform
    • SDK Design
    • System Design