I will build a custom ai chatbot on cloudflare workers with your own data


About this gig
Want a ChatGPT-style assistant on your own site or app - one that answers from YOUR content, runs on YOUR terms, and costs pennies instead of a $99/mo SaaS lock-in?
I deploy custom AI chatbots and backends on Cloudflare Workers - fast, secure, serverless, and cheap to run (free tier covers most small sites).
====================
WHAT YOU GET
====================
- A chat assistant embedded on your site (or as an API for your app)
- Answers from YOUR data - FAQ, docs, product info (Standard+)
- Runs on Cloudflare Workers: fast worldwide, scales to zero cost when idle
- Your own API key + provider (Claude / GPT / DeepSeek) - you own it all
- Hardened security: rate limiting, CORS, abuse protection - your API key never hits the browser (Premium)
- Clean README + walkthrough video included
====================
WHY ME
====================
I run this stack myself - a secure Cloudflare Worker proxy with a 10-layer defense setup (secrets, HMAC tokens, per-IP limits). You get hardened architecture, not a leaky demo exposing your API key.
Message me what the bot should do and where it will live. I will scope it and recommend the right package - free.
Get to know Jiaming
Custom HR Software, Employee Management, Inventory, Cloudflare and AI Automation
- FromChina
- Member sinceJul 2026
Languages
Chinese, English
My Portfolio
FAQ
Will my API key be safe? I have seen bots leak keys in the browser.
That is the #1 mistake I prevent. Your key lives only in the Cloudflare Worker (encrypted secret), never in the frontend. I add rate limiting and abuse protection so nobody can run up your bill.
Can it answer from my own documents or FAQ?
Yes - Standard and Premium feed your content into the assistant so it answers accurately about your product or business, not generic ChatGPT answers.
How much will it cost to run?
Cloudflare Workers has a generous free tier that covers most small and medium sites. You only pay your LLM API usage (often just a few dollars per month). No SaaS subscription.
Which LLM providers do you support?
Claude, OpenAI GPT, and DeepSeek. I will recommend the best price and quality for your use case.
How fast will you reply and start?
I reply within 3-4 hours, usually the same working day. You get my recommended approach and a clear price before any work starts. If your task needs more time or a different approach, I will say so straight away instead of going quiet.
Can I use a provider other than Claude or GPT - Gemini, Mistral, Llama or Ollama?
Yes. All of them work - the Worker just points at a different API endpoint and model name, and most of them speak the same OpenAI-style API, so switching is a one-line change. I will recommend the best price and quality for your use case.

