ChattyDocs

ChattyDocs organises work into datasets. You start by preparing one: upload multiple PDFs or any text file from your device with the built-in upload manager, or simply provide a website link and let the service download and parse the content for you. Documents can be added or removed at any time, and indexing happens in the background so you are not left waiting. A usage tracker shows how much of your upload credit or subscription quota each file or site will consume before you commit.

Once the dataset exists, you tune it. You pick a chat model and decide how much context, history and output space to allocate, adjust the temperature to control how creative or deterministic the answers are, and use system prompts to give the AI a name, a personality and a preferred language. Datasets can also be made public and shared by link, effectively turning your knowledge base into something other ChattyDocs users can query.

The chat experience is built around traceability and continuity. Answers arrive with their sources attached — file name plus page number for multi-page documents, or a direct link for websites — and you can export every question and answer with one click. Sessions are persisted, so you can pick up a conversation where you left off, and you can attach a standing chat context that gets folded into every query alongside the dataset content. The same sessions are visible whether you use the web app, an embedded web source or the Telegram bot, and ChattyDocs is available on desktop, mobile and Telegram.