Act as a Senior Frontend Developer and UI/UX Designer. Create a unique, professional, and eye-catching static portfolio website for "Akash", an AI Engineer, using only vanilla HTML5, modern CSS3, and vanilla JavaScript (ES6+). Do not use external CSS/JS frameworks like Bootstrap or React.

### 1. Design & Aesthetic Direction
- **Theme:** Futuristic Minimalism / Dark Mode Default (deep obsidian/slate background `#0a0e17`, electric cyan `#00f2fe` and neural violet `#7b2cbf` accents, high-contrast off-white text `#f0f6fc`). Include a working Dark/Light mode toggle.
- **Visual Style:** Bento-grid layouts, subtle glassmorphism (frosted glass cards with `backdrop-filter: blur`), glowing gradient borders on hover, and smooth micro-interactions.
- **Typography:** Pair a clean geometric sans-serif font (like 'Inter' or 'Space Grotesk' via Google Fonts) for headings/body with a monospace font (like 'JetBrains Mono' or 'Fira Code') for code snippets, model metrics, and tags.

### 2. User Experience (UX) & Performance Requirements
- **3-Second Clarity:** The hero section must immediately communicate who Akash is, what he builds, and provide clear calls-to-action (CTAs).
- **Interactive Background:** Implement a lightweight, high-performance HTML5 `<canvas>` neural network / particle node animation in the hero background that subtly responds to mouse movement without hurting scroll performance.
- **Scannable Hierarchy:** Use visual badges for tech stacks and bold metric callouts (e.g., "99.2% Accuracy", "40% Latency Reduction") on project cards so recruiters can scan achievements quickly.
- **Smooth Navigation:** Sticky glassmorphic navbar with active section highlighting on scroll, mobile hamburger menu, and smooth scrolling behavior.
- **Accessibility:** Semantic HTML tags, proper ARIA labels, keyboard-navigable interactive elements, and high text contrast.

### 3. Website Sections & Unique Features
1. **Hero Section:**
   - Status badge: "Available for AI/ML Roles & Consulting" with a pulsing green dot.
   - Headline: "Hi, I'm Akash — AI Engineer."
   - Dynamic JavaScript typing effect cycling through specialties: "Building Production LLMs", "Designing Computer Vision Pipelines", "Scaling MLOps Infrastructure", "Deploying Autonomous Agents".
   - Primary CTA ("Explore AI Projects") and Secondary CTA ("Download Resume" + quick social links for GitHub, LinkedIn, Hugging Face, and Kaggle).

2. **Interactive Terminal / "Ask Akash's AI" Widget (Unique UX Feature):**
   - A sleek, interactive command-line or chat-style card embedded on the page where visitors can click preset prompt chips (e.g., `whoami`, `top_skills`, `latest_model`, `contact_info`) or type simple commands, triggering simulated streaming text responses in vanilla JS.

3. **Technical Arsenal (Bento Grid):**
   - Visually categorized skill cards:
     - *LLMs & GenAI:* PyTorch, LangChain, LlamaIndex, RAG, Fine-Tuning (LoRA/QLoRA), OpenAI/Anthropic APIs.
     - *Machine Learning & CV:* TensorFlow, Scikit-Learn, OpenCV, YOLO, Transformers.
     - *MLOps & Backend:* Docker, Kubernetes, FastAPI, MLflow, AWS SageMaker, Vector DBs (Pinecone, Milvus).

4. **Featured AI Projects (Filterable Showcase):**
   - Include working JavaScript category filter buttons: `All`, `GenAI & LLMs`, `Computer Vision`, `MLOps`.
   - 4 detailed project cards displaying: Project Title, Architecture/Tech Stack tags, brief problem-solution summary, a highlighted performance metric badge, and links for "GitHub Repo" and "Live Demo / Hugging Face Space".

5. **Experience & Research Timeline:**
   - A vertical timeline highlighting AI engineering roles, model deployments, and research/publications with dates and quantifiable impact bullet points.

6. **Contact Section & Footer:**
   - Clean contact form with client-side JavaScript validation and simulated submission feedback, direct email copy-to-clipboard button with toast notification, and social links.

### 4. Code Deliverable Instructions
Provide the complete, production-ready, and well-commented code separated into three files:
1. `index.html` (complete semantic structure with realistic AI engineering placeholder data)
2. `styles.css` (CSS variables, responsive media queries, animations, and Bento grid layouts)
3. `script.js` (HTML5 canvas neural network animation, terminal widget logic, project filtering, scroll animations via IntersectionObserver, and copy-to-clipboard toast)

Do not use placeholders like `<!-- add more projects here -->` or truncate the CSS/JS. Write out the full implementation.