Deliverables

High-fidelity design
Assisting project management
Roadmap planning
Design system creation and documentation
Client presentations
Information architecture
User Flow Mapping
Assisting QA

🚀

Product launch

Fully launched the full scope product from mvp and beta testing versions, for the German citizens.

🔵

Consistent Performance

A flexible Agile process tailored to our company was adopted, ensuring high-impact solutions.

DIHK represents commercial and industrial enterprises and those belonging to the service sector vis-à-vis politicians, administrators and the public. It represents the general interest of the commercial sector at federal and European level – for example for less bureaucracy, free trade or a fast internet. The goal: good framework conditions for successful business. The DIHK aggregates the interests of businesses across all sectors and branches of the economy in a democratic and deliberative manner. These interests are conveyed to it by the 79 local Chambers of Commerce and Industry across Germany. The spectrum of opinions on the various economic policy issues is as diverse as the business landscape in Germany.

Create a single and unified platform for all digitised certified nationwide and non-nationwide seminars offered by a chamber of commerce entity through two main drivers:

• Innovation: using artificial intelligence;
• and Performance: providing all necessary information related to these seminars in the system as well as a booking management.

For DIHK to be able to get internal funding for the project, we quickly needed to come up with an MVP solution, that proved innovative and showed performance improvement. To tackle these requirements, we developed a chatbot search interface as the artificial intelligence component, which required the implementation of the Bot Training Management feature, to enable the training and optimising of the chatbot models search interface results, that would allow users to search for their desired seminar and trainings more efficiently and effectively.

The initial hypothesis was that users would search via a “main keyword” which we categorised as the “main topic” which defined how we built the bot sequence flows.

A little over a year later with the chatbot feature being continuously improved under a Beta Testing Version, me and the team finally received some important insights through user testing feedback on the chatbot sequences which directly affected the bot training feature in the back-office as well.

The main issue identified in the chatbot was:

• Unrelated suggested results based on the users keywords input requests

Whereas, the issues found in the bot training feature were:

• Uncertainty as to what keyword should be assigned as “main topic” and which could just be classified as “subtopics”
• Confusion and frustration towards the restriction on assigning one “main topic” keyword

User testing insights

From the insights and information gathered, we concluded that employees responsible for the tagging of offers in the back-office bot training management felt frustrated and confused by the current system for multiple reasons, but the main one being because we had created a chatbot model which required the user to provide information with a certain sequence (e.g. topic, date, location).

Through this sequence, we ended up restricting the user to always have to provide a “topic” keyword, which if not found or unrelated, would result in the process of failing at retrieving similar offer options when the specific keyword was not associated with multiple offers in the system.

Challenge

Iteration

Together with the rest of the team, we went through a series of workshops which I facilitated, analysing user-flows, use-case scenarios and sketching wireframes that allowed us to define and improve the functionalities that were not working, additionally and consequently improving the machine learning model, to optimise the bot training management, making it less restrictive with the keywords assignment on offers, and cancelling the previous negative influence retrieved in the chatbot search results.

Throughout this process, one thing we all agreed on that became very clear was that there was no real benefit (neither for the employees using the back-office system, or the users searching for offers in the chatbot) in the categorisation of keywords by “topic” and “subtopics”, so instead, we move forward by:

• No longer supporting the “topic” entity, and removing it from the systems model
• And also removing sub levels of keywords, and instead convert “topic” and “subtopic” keywords into general keywords

We then moved on to defining the new logic for this improved and optimised version of the bot training management system, the main goal being that every offer would now just have a set of keywords associated to it. The advantage of this system:

• Making the bot faster
• Removing the limitation of assigning a single keyword (topic) to a selected group of offers
• Allowing keywords allocation optional
• Eeasier to identify offers with no associated keywords

Moving on to the higher fidelity mockups, I made sure to design all detailed functionalities of the described use-cases.

The new bot training management called keywords assignment was much simpler and easier to use. By removing the layers of tagging of the offers with a “main topic” and “subtopic” keywords, and instead grouping the two elements into one, the user is less likely to assign a specific categorising “main topic” keyword to multiple offers, which the system previously assumed as the main keyword to retrieve results.

Statistic Insights

Having worked on the project since the very start, it has definitely given me a very deep understanding of the project, which when having to make these kinds of iterations which have a huge impact on the systems core functionality, I felt was really valuable and made it easier for me to iteratively find a solution that would meet the needs of the business and its customers.