Multi-Hub Business Matching web app with role-based access and an AI help center for a business network

How the brief came about

The challenge in this case was to bring the routine work of a multi-Hub business network into one system. That routine work covers meeting check-in, training, inviting guests, referring work to one another and giving sticker rewards. In 2026 we developed a Business Matching web app with 4 levels of user permissions for this business network group.

The customer wanted every Hub to use the same system, with members of each Hub mainly seeing their own Hub's data. Administrators at each level wanted an overview of the Hubs they look after, from meeting attendance through to the value of closed business. Members wanted quick answers to usage questions and did not want to repeat their story when passed on to an admin.

System requirements

The system had to serve every Hub in one place and separate what each user sees by role. The main requirements were as follows.

• Usable through a browser on both mobile phones and computers, including opening the camera to scan a QR Code for check-in.
• Members mainly see their own Hub's data; administrators see data within the scope they are responsible for.
• Sticker rewards must be transferred according to the business value recorded in the system, with no manual calculation.
• Redeeming stickers for cash must be approved by an admin before the balance is deducted.
• When the AI cannot answer a question, the admin must see the full conversation history, so the member does not have to explain again.

Why one system for every Hub, with permissions separated by role

We use one system for every Hub because members need to be able to search for members in other Hubs and transfer stickers to one another across Hubs. We therefore separate data visibility through permissions based on 4 role levels: member, Hub administrator, area administrator and system-wide administrator.

The system is a web application (used through a browser, with no app to install), so members can use it straight away from a mobile phone or computer. For work involving money and answers the AI does not know, we made people the decision-makers: an admin approves cash redemption, and an admin continues the chat from the AI.

What does this system consist of?

The main parts of this system are QR Code check-in, asking for help and referring work, recording business value, the Sticker Wallet, the AI help centre, and dashboards and reports. These parts connect from meeting attendance through to the administrators' reports. Each user sees the menus and data for their own role.

1. 1

   QR Code meeting check-in

   An administrator creates the Hub's meeting, then prints the entry and exit QR Codes and places them at the venue. Members open their mobile phone camera and scan to sign in. A member who cannot attend can register a substitute in advance.

2. 2

   Asking for help and referring work

   A member posts what they need on The Ask Board (a board for requesting help). Someone who can help accepts, and the system then takes them to create a Job Recommend to refer the work, specifying the customer's name and the level of urgency.

3. 3

   Record the value of closed business

   When a deal closes, the person who received the work records its value on the Success Business page. The system changes the job status to successful and automatically transfers gold stickers to the person who referred the work, all in one step.

4. 4

   Use and redeem stickers

   The Sticker Wallet shows the balance of each sticker type, which members can transfer to one another or use to pay for training. A cash redemption request shows the net amount before confirmation and requires admin approval before the balance is deducted.

5. 5

   Ask the help centre

   A member selects a topic and types a question. The AI Chatbot answers according to the selected topic and the user's role. If it cannot answer, the system says it is connecting to a staff member, then passes the chat room with its history to an admin to continue.

6. 6

   View the overview through dashboards and reports

   Hub administrators and area administrators can view statistics on meeting attendance, work referrals and business value for the Hubs they are responsible for, from the dashboard and the reports page.

Besides this main process, the system also has registration for guests invited to meetings by members, and real-time chat between users.

Development process and request-based improvements

We developed this system in rounds, starting with the core functions members use at every meeting, then continuing to adjust it according to user requests after they had tried it.

• Core functions already working: QR Code check-in, work referral with automatic sticker transfer, training, and a help centre where the AI hands over to an admin.
• User requests already implemented: role-based dashboard filters, a reports page and email notifications.
• Other user requests: planning for the next round is in progress.

Confirmed results so far

At this point, the system's core functions are working and further improvements in response to user requests are ongoing. When the person who received the work records the value in Success Business, the system transfers gold stickers to the person who referred the work in the same step.

Requests submitted by users and already implemented include role-based dashboard filters, a reports page and email notifications. There are no verifiable usage figures yet, because the project has not closed.

What we learned

This work offers the following principles for member systems with multiple groups and rewards.

• Tie rewards to entries recorded in the system, in the same step. When recording the value and transferring stickers happen together, the wallet balance matches the recorded business value and no manual calculation is needed.
• Define the points where AI hands over to a person, and the decisions people must make themselves. In this work, the AI answers questions by topic first, questions it cannot answer go to an admin with the history, and every cash redemption requires admin approval.
• Design visibility permissions together with dashboards and reports, because administrators at each level need to see different data. In this work, the requests implemented after users had tried the system include role-based dashboard filters and a reports page.

How can you apply this approach to your network?

If your organisation has several member groups that meet regularly and refer work to one another, you can use the same approach as this case: one system with role-based permissions, with rewards tied to recorded amounts. The parts to adjust are the reward rules, the administrator roles and the questions the AI must answer.

• Define clearly from the start who can see data across groups and who can approve transactions.
• Write the points, sticker and cash redemption rules down before development, because these rules determine the calculation formulas and reports.
• Prepare the questions members often ask, with correct answers, and decide who takes over when the AI cannot answer.
• Allow time for improvement rounds based on requests, because users often ask for changes to screens and reports once they have tried the system.
• If you want to accept online payments or send SMS, plan the integration with external providers as separate work.