Product overview

How does the platform work?

Upload your documents, build AI assistants, and answer customers through API or the built-in test chat. Each business account has its own isolated data. You choose a plan, pay online, and start immediately.

From upload to answer

The journey in five simple steps:

πŸ“„

Upload files

PDFs and text in your knowledge library

βš™οΈ

Processing

Files are indexed automatically

πŸ€–

Assistant

You configure tone, model and tools

πŸ”

Smart search

Relevant content is found for each question

🌐

Answer

Via API or test chat

πŸ–₯️

Admin dashboard

Everything in one place

After signing in you can:

  • βœ“ Create knowledge libraries and upload PDF or text files (up to 1 GB each)
  • βœ“ Build assistants: name, instructions, AI model, attached libraries
  • βœ“ Track file processing status in the embedding queue
  • βœ“ Test conversations in the playground before going live
  • βœ“ Connect live database tools for orders, profiles, inventory
  • βœ“ Manage wallet, plans and usage
πŸ“š

Knowledge libraries

Your files become searchable knowledge. When a customer asks a question, the assistant finds the most relevant parts of your documents and answers based on that content β€” not guesses. Steps: 1. Create a library and upload files 2. Attach the library to an assistant 3. Wait until processing shows β€œready” 4. Ask questions via test chat or API

  1. 1 Create library β†’ upload files
  2. 2 Attach library to assistant
  3. 3 Wait until status is ready
  4. 4 Go live with API or test chat

You only see libraries belonging to your account. Admins see all accounts.

πŸ€–

AI assistants

Each assistant has its own personality, AI model, knowledge libraries and optional database tools. You receive a private API token to integrate with your website or app. Status meanings: β€’ Processing β€” files are still being indexed β€’ Active β€” ready to answer β€’ Inactive β€” manually paused

  • βœ“ Instructions: role, language, rules, when to use database tools
  • βœ“ Choose the AI model that fits your budget and quality needs
  • βœ“ Attach only the libraries that assistant should use
πŸ”Œ

API integration

Send customer messages to your assistant with a simple HTTP request and Bearer token. Ideal for websites, mobile apps and support widgets. See the API guide for copy-paste examples.

Usage is metered according to your plan.

πŸ—„οΈ

Live database answers

When the answer is not in a document β€” for example β€œwhere is my order?” β€” the assistant can query your MySQL database safely using predefined read-only queries you configure. Example: order status, account details, stock levels.

  1. 1 Connect your database in assistant settings
  2. 2 Define named queries with clear descriptions
  3. 3 Mention in instructions when each query should be used
πŸ’¬

Conversation memory

Optional feature: the assistant remembers recent messages in the same conversation so follow-up questions make sense. You control how many previous turns are kept.

πŸ’³

Plans and billing

Choose a plan on the pricing page, pay online, and your account limits update immediately. Plans control number of assistants, libraries, storage, monthly messages and premium features.

πŸ”’

Security and privacy

β€’ Each account sees only its own data β€’ Keep API tokens secret β€’ Use read-only database users for live queries β€’ Super admins manage the whole platform

Ready to choose a plan?

See pricing on the home page, pick the plan that fits you, and start building your first assistant.