We allow users to build web apps and chatbots by chaining LLM APIs (we call processors) from providers like OpenAI, StabilityAI, Cohere etc, without writing any code. We also let users to bring in their own data and store it in a vector database to be used for context augmentation in their apps. Users can import data from a variety of sources including urls, sitemaps, PDFs and other file types

Published apps are accessible to app's users via a public url. Chat apps can also be embedded as a chat bubble on external websites. There are also high level processors offered by Promptly which unlocks things like extracting text from files, youtube urls etc., enabling users to build some sophisticated applications.

Here are some example applications: Resume generator - https://trypromptly.com/app/d478594d-2082-46c9-bee7-f057f4bc... - Uses single ChatGPT processor to generate tailored resume based on user and job information

Story generator - https://trypromptly.com/app/9d6da897-67cf-4887-94ec-afd4b936... - Takes in a theme, generates a short story and an illustration to go with it. This is built with three processors, ChatGPT to generate story -> ChatGPT to generate DALL-E prompt -> DALL-E image generation.

Promptly support - Chatbot on Promptly website itself is built on the platform with website url as the datasource. https://twitter.com/ajhai/status/1643521969072242689 shows how the chatbot was built.

We built Promptly as a prompt and LLM workflow management platform out of our experience using LLMs at MakerDojo and landed here. Please try it out and let me know if you have any feedback or questions.