Nikhil Bhagat

Senior Backend Developer | Full Stack Engineer | AI Integration Specialist

8+ years building scalable APIs, real-time systems, and AI-powered products
Specialized in Python backends, cloud-native architectures, and Generative AI integration

Full Stack Engineer with 8+ years of hands-on experience designing and deploying production-grade backend systems, RESTful APIs, and real-time applications. Currently working as Senior Backend Developer at CineART (Hong Kong) since June 2022, leading backend architecture for scalable social platforms.

I specialize in Python-based backends (Django, FastAPI), cloud infrastructure (AWS, Docker, Kubernetes), and integrating Generative AI solutions using OpenAI API, LangChain, and custom transformer models. My work focuses on system design, performance optimization, and building intelligent automation-first solutions that handle real-world scale.

Senior Backend Developer | CineART • Hong Kong June 2022 – Present

• Leading backend development for a LinkedIn-style professional networking platform targeting Nepal market
• Architecting microservices using Django REST Framework, Docker, and planning AWS cloud migration
• Developed engagement-based ranking algorithm to optimize content visibility and user interaction
• Implemented secure JWT authentication, Redis caching, and performance optimization strategies
• Set up CI/CD pipelines with GitHub Actions and integrated third-party service APIs
• Mentoring junior developers on test-driven development and code quality best practices
• Collaborating with cross-functional teams to deliver features on tight deadlines

Languages & Frameworks Python JavaScript (ES6+) SQL TypeScript Django Django REST Framework FastAPI Node.js Express.js React.js Next.js Redux TailwindCSS

AI & Generative AI OpenAI API LangChain LlamaIndex Transformers PyTorch Vector Databases (FAISS, Pinecone) Fine-tuning (LoRA) NLP

Databases PostgreSQL PostGIS MongoDB Redis

Cloud & DevOps Docker Kubernetes AWS (Lambda, EC2, S3) Heroku GitHub Actions Jenkins Terraform Ansible

Architecture & Patterns RESTful APIs Microservices Event-Driven Design Kafka Webhooks Background Workers (Celery + Redis) WebSockets Firebase

Testing & Monitoring Pytest Postman Django Debug Toolbar Logging & Metrics

Jul 2024 – Present | Private Repository

Building a custom Large Language Model from scratch, fine-tuned for code understanding, generation, and debugging assistance.

• Architecture: Implemented multi-head self-attention, positional encoding, layer normalization, and causal masking using PyTorch
• Training: Pre-trained on 50K+ curated code files (Python, JavaScript, TypeScript) for syntax understanding
• Optimization: Applied LoRA fine-tuning and int8 quantization for efficient deployment on limited GPU resources
• Features: Function summarization, error explanation, and unit test generation capabilities
• Deployment: Real-time inference via FastAPI microservices integrated with LangChain pipelines
• Challenges Solved: Gradient explosion and overfitting using gradient clipping and adaptive learning rate scheduling

Tech Stack: PyTorch Transformers FastAPI LangChain CUDA Tokenizers LoRA

Aug 2024 – June 2025 | Client Project (Japan)

Built a real-time backend for a location-based multiplayer game where users claim zones, engage in strategic attacks, and compete for territory ownership.

• Core Features: Geofencing, real-time push notifications, live synchronization, zone battles, and official leaderboard (top 100 players)
• Architecture: RESTful APIs with Django REST Framework and GeoDjango for spatial operations
• Database: PostgreSQL + PostGIS for spatial data queries and territory calculations
• Background Processing: Celery + Redis for async task handling
• Real-time Updates: Firebase Cloud Messaging for push notifications
• Security: JWT authentication with role-based access control

Tech Stack: Django DRF GeoDjango PostgreSQL PostGIS Redis Celery Firebase

Mar 2025 – Oct 2025 | Client Project (Nepal)

Developed an intelligent document-driven conversational AI system to assist with Nepal Financial Reporting Standards (NFRS) in English and Nepali.

• AI Capabilities: Semantic document processing and vector-based retrieval using FAISS for precise financial query responses
• Authentication: Secure JWT-based user authentication system
• Multilingual: Full support for English and Nepali language queries
• Performance: Context-aware guidance with real-time response generation
• NLP Integration: Advanced language models for understanding complex financial terminology

Tech Stack: Django REST Framework FAISS LangChain JWT PostgreSQL NLP Models

Jan 2024 – Mar 2024

Built an AI-powered chatbot to analyze and explain large codebases, reducing debugging time by 30%.

• AI Integration: OpenAI API and LangChain for code analysis and natural language explanations
• Features: Prompt templates, error tracing, code-to-English explanations
• Optimization: Solved token limit issues for large files through context compression
• Impact: Helped developers understand complex codebases faster

Tech Stack: Python LangChain OpenAI API Django REST Framework

Sep 2022 – Oct 2023 | Client Project (Hong Kong)

Developed a scalable backend for a TikTok-style community app supporting 10,000+ users with real-time posts, likes, and comments.

• Real-time Features: WebSocket-based notifications for instant user engagement
• Media Handling: Cloudinary integration with optimized video caching via stream chunking
• Analytics: MongoDB aggregation pipeline for user behavior insights
• Performance: Solved upload timeouts and optimized API response times

Tech Stack: Node.js Express.js MongoDB Cloudinary WebSocket

Feb 2025 – Aug 2025 | Client Project (Philippines)

Built a Trello-style Kanban board with real-time collaboration features for task assignment and progress tracking.

• Features: Drag-and-drop interface, role-based authentication, activity logs
• Real-time Updates: Socket.IO for live task movements across users
• Reliability: Implemented heartbeat checks to resolve WebSocket disconnect issues
• UI/UX: React Beautiful DnD for smooth drag-and-drop experience

Tech Stack: React.js Node.js Express.js Socket.IO PostgreSQL

Feb 2020 – June 2022 | Client Project (Italy)

Created a full-stack platform for developers to share portfolios and collaborate on projects.

• Frontend: React with JWT authentication, Redux state management, and Axios API integration
• Backend: REST APIs in Express.js with MongoDB database
• Media Storage: Cloudinary for image uploads and CDN delivery
• CI/CD: GitHub Actions for automated testing and deployment
• Performance: Improved build performance by 40% and resolved CORS and API throttling issues

Tech Stack: React.js Node.js Express.js MongoDB JWT Cloudinary GitHub Actions

Mar 2024 – Jan 2025 | Open Source

Automated Node.js microservice scaffolding tool, cutting setup time by 50%.

• Installation: npm i turbogen (1,000+ weekly downloads)
• Features: Automatic project structure, dependency installation, boilerplate generation
• UX: Enhanced CLI experience using Commander.js and Chalk for colored output
• Bug Fixes: Resolved recursive dependency installation issues

Tech Stack: Node.js CLI Tools Commander.js Chalk

Jul 2018 – Dec 2018

Created an interactive visualization tool demonstrating how neural networks propagate data.

• Visualization: D3.js for rendering neural network nodes and connections
• Performance: Implemented virtualized nodes to handle large networks without lag
• Impact: Adopted by 200+ learners for educational purposes
• Educational Value: Simplified complex AI concepts through visual representation

Tech Stack: React.js D3.js TensorFlow

Bachelor of Computer Application (BCA) Mahendra Morang Aadarsha Multiple Campus | 2025

Secondary School Certificate Merryland College | 2013

• npm Publisher: Created and maintain Turbogen (1,000+ weekly downloads) and 4+ utility packages
• Open Source Contributor: 75+ public repositories with 50+ stars on GitHub
• Active Contributor: Contributing to AI/ML tools and developer productivity projects
• Community Impact: Tools and projects used by hundreds of developers globally

• API Design & Development
• System Architecture & Scalability
• Performance Optimization
• CI/CD Pipeline Implementation
• Agile Development Methodology
• Problem-Solving & Debugging
• Cross-functional Collaboration
• Technical Documentation
• Code Review & Mentorship

experience = {
    "years": 8,
    "role": "Senior Backend Developer",
    "specialization": [
        "RESTful API Design",
        "Real-time Systems",
        "AI Integration (OpenAI, LangChain)",
        "Microservices Architecture",
        "Cloud Infrastructure (AWS, Docker, Kubernetes)"
    ],
    "current_focus": [
        "Generative AI Integration",
        "Transformer Model Development",
        "Cloud-Native Architecture",
        "Performance Optimization"
    ]
}

• Generative AI & Large Language Models
• Cloud-Native Architecture & DevOps
• System Design & Performance Engineering
• Open Source Development & Community Building