const devan = {
    pronouns: "He/Him",
    location: "Nadiad, Gujarat, India 🇮🇳",
    education: "B.Tech CE @ Dharmsinh Desai University",
    currentCPI: "8.39",
    currentFocus: "Full Stack Development & AI/ML",
    askMeAbout: ["Web Dev", "Applied AI", "Problem Solving"],
    technologies: {
        frontEnd: ["Next.js", "React", "TypeScript", "Three.js"],
        backEnd: ["Spring Boot", "Node.js", "Express.js", "Fastapi"],
        ai: ["LLM", "GenAI", "RAG"],
        databases: ["PostgreSQL", "MongoDB", "MySQL", "Redis"],
        cloud: ["Docker", "Vercel", "Railway", "AWS"]
    },
    funFact: "Won Consolation Prize at Holbox 2025 hackathon! 🏆"
};

Tech Stack: Next.js, TypeScript, Clerk, Java, Spring Boot, PostgreSQL, Docker, Traefik, Redis, Consul, API Gateway, AWS, GitHub Actions

• 🧩 Built a backend engineering practice platform where developers solve real-world challenges with isolated code execution, database integration, and automated evaluation.
• 🐳 Engineered a secure execution engine using Docker sandboxing, Redis queues, and distributed workers for asynchronous code execution and scalable task processing.
• ⚙️ Designed a Spring Boot microservices architecture with Consul service discovery, API Gateway routing, centralized authentication, rate limiting, and WebSocket-based real-time log streaming.
• ☁️ Deployed and operated containerized services using AWS and Traefik, with automated deployments through GitHub Actions and production-oriented traffic management.

Tech Stack: Next.js, Express.js, TypeScript, MongoDB, Redis, BullMQ, Cloudflare R2, Hugging Face, Razorpay

• 🎬 Developed a full-stack AI video generation platform supporting script-to-video creation, AI voice synthesis, payment-based credits, and social content sharing.
• ⚡ Designed an event-driven backend using Redis, BullMQ, SSE, and webhooks for long-running AI jobs, real-time progress tracking, and payment lifecycle events.
• ☁️ Built a production-ready media pipeline with Cloudflare R2 and CDN delivery for scalable video and asset storage.
• 🔐 Implemented secure authentication, rate limiting, optimized asset management, and Razorpay-based credit payments.

Tech Stack: Next.js, FastAPI, Python, TypeScript, WebSockets, Deepgram, Groq, Edge-TTS, Three.js, Agentic AI

• 🎙️ Developed a real-time multimodal AI agent combining speech recognition, LLM reasoning, tool calling, and voice synthesis to perform user-requested actions.
• ⚡ Architected a streaming voice pipeline using Deepgram STT, WebSockets, GPT-OSS-120B, and Edge-TTS for context-aware, low-latency conversations.
• 🧠 Implemented an agent runtime with tool orchestration and state management for autonomous multi-step actions.
• 🔌 Integrated tools including Spotify, YouTube, WhatsApp, Google Meet, weather services, and browser automation.

• 🚀 Building: AI-powered solutions combining LLMs with web technologies
• 📚 Learning: Advanced backend development and ML based solution
• 🌱 Exploring: LLM Learning
• 🤝 Contributing: More to open source projects and community initiatives
• 🎨 Creating: Interactive 3D web experiences with Three.js