$ Harishkumar Pillai
$ ai-chat

Ask My LLM

Chat with an LLM that knows about my projects, experience, and technical work. Powered by Cloudflare Workers AI with guardrails and grounding.

✓ Grounded

Answers only from my curated knowledge base. No hallucinations about my work.

✓ Guarded

Llama Guard 3 moderates input/output. Refuses NSFW, illegal, or prompt injection attempts.

✓ Cheap

Low-cost serverless inference (usage-based). No GPU hosting costs.

ai-chat-terminal

Welcome to the AI chat! Try asking:

  • ▸ "What DNS security projects have you worked on?"
  • ▸ "Tell me about your LLMOps experience"
  • ▸ "What's your tech stack?"
$
⚠️
Implementation Status:

Chat backend is being deployed to Cloudflare Workers. The worker will handle moderation, KB retrieval, LLM inference, and rate limiting. Check back soon!

How It Works

  1. Your message is sent to /api/chat (Cloudflare Worker)
  2. Llama Guard 3 moderates input for safety (rejects NSFW, injections, etc.)
  3. KB chunks are retrieved based on keywords and tags
  4. Llama 3 generates answer using ONLY the provided context
  5. Output is moderated again before being returned
  6. Rate limit: 30 requests/day per IP (stored in Cloudflare KV)