I'm a Senior Agentic AI Engineer and Backend Architect from Nepal, building production-grade AI systems that actually ship and scale.
Over 6+ years, I've gone from full-stack development to specializing in the frontier of AI — autonomous multi-agent systems, LLM orchestration pipelines, RAG architectures, and high-performance FastAPI backends. My current focus is at the intersection of AI and real-world product engineering.
Right now I'm leading AI development at KLIKY AI / WinstaAI — a production Agentic AI SaaS platform — orchestrating GPT-4o, Claude 3, Gemini, ComfyUI, and LangGraph multi-agent workflows at scale.
Open to: Senior AI Engineer · LLM Engineer · Agentic AI Developer · Backend Architect · Full-Stack AI roles — Remote, Contract, Freelance, Staff Augmentation — USA · UK · UAE · Europe · APAC · Worldwide
|
🤖 AI Platforms Multi-agent SaaS products with LangGraph, CrewAI, and LLM orchestration |
🔍 RAG Pipelines Document intelligence with LlamaIndex, Pinecone, Qdrant, ChromaDB |
⚡ FastAPI Backends Async microservices, event-driven systems, Kafka, WebSockets |
🎨 Generative AI Text-to-image, face swap, speech AI, video dubbing, OCR workflows |
🤖 Agentic AI & LLM Systems ████████████████████ LangGraph · CrewAI · AutoGen · Tool-Calling
🔍 RAG Pipeline Architecture ███████████████████░ LlamaIndex · Pinecone · Qdrant · ChromaDB
⚡ FastAPI Microservices ████████████████████ Async · Containerized · Cloud-Native
🧠 LLM Orchestration ██████████████████░░ GPT-4o · Claude 3 · Gemini · DeepSeek
🎨 Generative AI █████████████████░░░ Image · Video · Audio · Diffusion Workflows
🌐 Event-Driven Systems ████████████████░░░░ Kafka · Debezium CDC · Celery · Redis
☁️ Cloud & GPU Inference ██████████████░░░░░░ RunPod · FAL.ai · AWS · GCP · K8s
|
LangGraph · CrewAI · GPT-4o · Claude · Gemini · FastAPI · Qdrant Production AI SaaS platform with LangGraph multi-agent workflows, intelligent LLM routing across GPT-4o / Claude 3 / Gemini, text-to-image (ComfyUI + Replicate), face swap, OCR, video/audio AI, FastAPI microservices, GPU inference (RunPod + FAL.ai), and a full credit/billing system.
|
Django · Speech AI · DeepL · Firebase · REST API AI-powered multilingual platform with speech-to-text, text-to-speech, real-time transcription, and video dubbing. Served global audiences via Generative AI services, secure REST APIs, and third-party translation integrations (DeepL).
|
|
LLM · GPT-4 · Generative AI · Firebase AI-based tool that generates structured presentations and roadmaps from unstructured text and images. Full-cycle: LLM content generation pipelines, Firebase auth, OTP, push notifications, and REST APIs.
|
Django · PostgreSQL · RBAC · Docker HIPAA-compliant medical diagnostic reporting platform serving a large hospital network. Real-time diagnostic imaging, secure role-based access control, scalable reporting modules, and optimized healthcare workflows.
|
Enterprise-grade, production-ready starter templates — built to deploy, not just demo.
|
🐍 FastAPI Backend JWT · RBAC · PostgreSQL · Docker · K8s |
🟢 Node.js Backend Express · Prisma · PostgreSQL · Docker |
🦄 Django Backend DRF · Auth · Admin · REST · PostgreSQL |
|
⚫ Next.js Frontend TypeScript · Tailwind · App Router |
⚛️ React Frontend Vite · TypeScript · React Router |
📦 Explore All 7+ actively maintained repos |
| Period | Role | Company | Highlights |
|---|---|---|---|
| Jan 2024 – Present | Lead AI Full-Stack Developer | KLIKY AI / WinstaAI | LangGraph multi-agent platform · GPT-4o/Claude/Gemini routing · FastAPI microservices · GPU inference (RunPod, FAL.ai) · Stripe/PayPal billing |
| Jun 2021 – Dec 2023 | Full-Stack Developer | AJATH INFO TECH | Dibbly multilingual video AI · Speech-to-text/TTS · DeepL translation · Django REST · Firebase |
| Dec 2019 – May 2021 | Full-Stack Developer | SARAJ SYSTEM | PathHub HIPAA medical platform · Khantailor eCommerce · OZMED SaaS LMS |
- WinstaAI — shipping new LangGraph agentic workflows with autonomous decision-making and memory
- Production RAG engines — hybrid search with LlamaIndex + Qdrant reranking pipelines
- Agentic automation systems — CrewAI / AutoGen with persistent tool-calling and multi-step execution
- Open source backend templates — enterprise FastAPI, Django, and Node.js starters at fullstack-open-source
Available for Remote Hire · Worldwide