GRAALSOFT PROJECT

AI chatbot for Azitis.com

Auction questions, answered through conversation.

Client: Azitis
Solution: AI chatbots
GraalSoft contribution: Development and improvement of the conversational AI component.

DOCUMENTED RESULT · SEPTEMBER 2026

124 signed-in users turned to chat in one month.

1–30 September 2026 · Azitis internal usage report

Azitis has an in-platform assistance channel that users turn to for questions about assets, participation and account access. The 124 accounts started 208 conversations; another 68 conversations came from signed-out visitors.

124distinct signed-in users

276recorded conversations

573user messages

Source: internal report, 1–30 September 2026. Totals include launch and testing activity. The report does not include total site traffic or deduplicate signed-out visitors, so it cannot establish the share of visitors using chat. Conversation counts do not represent resolved requests.

01 / THE NEED

Where the project started

Azitis needed assistance at the points where users search for an asset, try to join an auction or need help with the platform’s steps. Information also needed to be accessible through an everyday-language question.

02 / THE DEVELOPMENT

What we built for this project

We developed and improved the AI conversational layer of Florin, the assistant integrated into Azitis.com. Work covered connecting conversations to platform information, instructions guiding responses and assessing behaviour against questions from actual use.

AI chatbot interface showing the assistant introduction, conversation area and message field. — IN THE AZITIS INTERFACE
Help starts where the question arises.
The assistant is accessible within the platform. Users type their questions in the same interface where they explore auctions. The screenshot shows the conversation’s opening screen.

Knowledge connected to conversation

Organising information that supports answers so that the assistant can use product context.

Answer improvements

Working on instructions and how available information is used in the dialogue.

Integration and evaluation

Connecting the AI component to the application and checking behaviour on questions relevant to the platform.

03 / CLIENT VALUE

What the development brings to the product

An assistance channel inside the product

Users can ask for explanations within the platform, at the moment a question arises. Conversation provides an entry point to auction and participation information.

Visibility into customer needs

Questions reveal what people seek, where they get stuck and what needs clarification. For Azitis, conversations also provide concrete input for prioritising improvements.