This product is specifically designed for Kwinzo B.V, it helps the company analyze policy documents, compare them with selected ISO norms and annexes, identify whether requirements are fully met, partially met, or missing, and generate compliance checklists, which improves the efficiency of their compliance assessment process.
Kwinzo B.V. is a company that helps organizations obtain certifications for information security and quality standards. We held regular in-person and online meetings with the client to confirm project details and also provided updates via email.
We had 7 members with clearly defined roles, including a main product owner, a main scrum master, and a main tech lead. At the beginning of the project we adopted the Scrum methodology and planned our work over 8 two-week sprints. In the first phase of the project, we all worked together to research the feasibility and define the scope of the project. After completing the research phase, we entered the implementation phase, where we divided the team into three subgroups: UI, API, and NLP, and this structure allowed each team to focus on their specific tasks and start working efficiently.
We also assigned secondary and tertiary roles, such as a secondary product owner, a secondary scrum master, and additional tech leads to support the main roles. This helped ensure smooth communication across subgroups.
One of the main challenges we faced was that the client had high expectations regarding AI, and the exact requirements were initially quite vague. This made it difficult to define a clear direction at the start for our team. However, through communication, meetings with the client, and careful research, we were able to better understand the client's needs as well as our capabilities. As a team we worked closely with the client to clarify priorities and identify realistic, manageable solutions for his company needs. In the end, this approach allowed us to deliver our final product.
Another challenge was maintaining steady progress while taking care of multiple aspects of the project. To overcome this, we split into smaller subteams, each focused on a specific area of our work. This allowed us to work in parallel and stay on track.
We're proud of our project, even if it wasn't clear at the start. Over time, we made strong, consistent progress which resulted in an AI and NLP assistant for our client's company. We integrated support for ISO standard updates, built a clear and intuitive UI, and made sure that our API model is capable of processing many files at once.