About This App
I (Jeremy Bloom) built this LLM Demo App to demonstrate three things:
- How easy it is to prototype with large language models (LLMs) in Python
- A hands-on way to brush up on my AWS and Python skills
- And (hopefully) to convince clients or employers that I do more than just build slide decks
This app lets you experiment with several LLM services:
- ChatGPT via OpenAI’s
openai
Python module
- Claude via Anthropic’s
anthropic
Python module
- Titan via AWS Bedrock
- …?
You can run the app in two modes:
server_keys
– Keys are provided by me (with built-in rate limits)
user_keys
– You provide your own API keys (removes rate limits for ChatGPT & Claude, not avail for AWS)
Caveats
- This is a prototype, likely running on a development server - it's not hardened or secured for production.
- While the app doesn’t store any keys, prompts, or responses, they may be visible in logs, debugging sessions, or edge-case errors.
- The source code is here:
github.com/databloomnet/demo1
Things you can do...
Logs:
Mode: server_keys
Try OpenAI API:
Try Anthropic API:
Try AWS Bedrock Titan: