Full-Stack Developer · AI/ML Systems · Competitive Programmer
Building production systems across web, mobile, and ML infrastructure. Expertise in Node.js, Python, React, real-time architectures, and AI integration. Currently exploring quantitative systems and agentic AI. Based in Pokhara, Nepal.
Backend & System Design Real-time architectures, REST/GraphQL APIs, microservices, database optimization, async job processing. Node.js, Express, tRPC, FastAPI.
Frontend & Mobile Responsive web UIs and cross-platform mobile apps. React, React Native, Expo. Production-grade state management and performance optimization.
AI & ML Systems LLM orchestration, vision APIs, prompt engineering, multi-model inference pipelines. Integration with NVIDIA NIM, Claude, Gemini. Building practical AI agents and automation.
Competitive Programming & Algorithms Strong DSA foundation. Active on LeetCode — focusing on system design patterns and optimization techniques applicable to real-world engineering.
Quantitative research system for the Nepal Stock Exchange with real-time data pipelines and ML-driven signal generation.
Problem Solved: Fragmented data sources, lack of institutional-grade analytics for NEPSE.
- PostgreSQL backend with Docker deployment
- Real-time data ingestion and validation pipelines
- Feature engineering for LSTM signal generation
- Modular architecture for multi-strategy backtesting
Tech Stack: Python (Pandas, NumPy, scikit-learn), PostgreSQL, Docker, PyPI package integration, Feature Store design
Enterprise helpdesk SaaS platform. Multi-tenant ticket management with real-time live chat, role-based access control, and Stripe billing integration.
Key Features: Queue-based ticket routing, WebSocket-powered live chat, knowledge base self-service, audit logging.
Tech Stack: React, Node.js/Express, PostgreSQL, Socket.io, Stripe, JWT authentication
Intelligent tutoring app built for mobile and web. Processes document images in real-time, provides AI-powered study guidance, and adapts to learning patterns.
Key Features: OCR + vision API integration, OAuth authentication, responsive React Native UI, scalable Express backend.
Tech Stack: React Native/Expo, Node.js/Express, Google Cloud Vision API, MySQL, OAuth 2.0
Full-stack AI learning companion with modular ML microservices. Integrates multiple LLM providers (NVIDIA NIM, Claude, Gemini) with intelligent fallback routing.
Key Features: Provider-agnostic LLM abstraction, FastAPI microservice, React Native frontend, structured recommendations.
Tech Stack: React Native, Express, FastAPI, NVIDIA NIM, MySQL, Drizzle ORM, TypeScript
Telegram-integrated agentic AI system with multi-tool orchestration and persistent state management. Built with Gemini 2.5 Flash.
Key Features: Tool composition, conversation memory, dashboard monitoring, Telegram bot integration.
Tech Stack: Node.js, Gemini API, Telegram Bot API, local persistent storage
Languages: Python · JavaScript · TypeScript · C · SQL
Frameworks & Libraries
- Frontend: React, React Native, Expo, Tailwind CSS, shadcn/ui
- Backend: Node.js, Express, tRPC, FastAPI
- Databases: PostgreSQL, MySQL, MongoDB
- DevOps: Docker, Docker Compose, Git
AI/ML & Integrations
- LLM Orchestration: Claude, NVIDIA NIM, Google Gemini
- Vision APIs: Google Cloud Vision
- Agentic Systems & Prompt Engineering
- Data Processing: Pandas, NumPy, scikit-learn
📧 Email: poudelashish572@gmail.com
🐙 GitHub: @Aashish-po
🧠 LeetCode: aashish___07
💼 LinkedIn: Aashish Paudel
- First-semester BSc CSIT at Tribhuvan University, Pokhara — balancing academics with independent project work
- Deep interest in: Distributed systems, real-time architectures, quantitative systems, AI integration patterns
- Problem-solving approach: Ship working code first, then optimize and scale
- Always: Learning new tools, shipping projects, solving hard problems