Challenge 1: Complicated certification requirements
The process for obtaining ISO certification includes documentation reviewing, auditing, process correction and improvement, reassessing, etc, before a certification can be issued. Typically, after the auditors visit and review companies’ management systems on-site, they write reports about the deviations, according to the certification clauses and requirements, and corrective recommendations. However, each ISO standard comprises many requirements. For example, ISO 9001:2015 Quality management system consists of 10 clauses and more than 350 requirements. It was absolutely time- and energy-consuming for the auditors to go through each of the underlying requirements, decreasing the time spent on the real, meaningful interaction with the entities to be inspected.

Challenge 2: Auditors’ efficiency and writing consistency
While conducting audits, the auditors have to not only review the documentation but also interact with the clients, ask them questions, and examine their management systems. Ideally, Qvalify prefers the auditors to spend more time with the clients to get a deeper understanding of their operations. With that said, seeking an effective way to reduce the time spent on writing reports would be a perfect move. In addition, having around 55 auditors, Qvalify had struggled to keep the works of each auditor consistent. Both aligning the style of writing to a certain standard and managing the quality of the reports were of Qvalify’s top priority. Apart from internal training and education, Qvalify was keen to find intelligent solutions to assist the auditors when they write the reports.

Challenge 3: Make good use of the historical data
Making proper use of the abundance of historical data was a great challenge. In their reporting system, Qvalify has kept around 20 years of historical data, which are expected to be used as handy references while doing audits. Like many other companies, Qvalify seeks to exploit the data and maximize the use of them. Instead of trying to build their own AI models, Qvalify took a more efficient and less costly approach: seeking a professional solution provider.

Through some word-of-mouth recommendations, Qvalify found Labelf and worked closely with us to tackle the challenges. After having a few meetings and getting some of the data, Labelf came up with solutions for Qvalify’s business and seamlessly built the AI model into Qvalify’s reporting system Quno.

Structuring and labeling the data contribute to the logical use of the existing reports. With the new application programming interface (API), the AI models find the corresponding criteria for ISO as soon as the auditors type in the reporting system. Rather than selecting the standard fields manually and scanning through the requirements one by one, now the system can automatically suggest what are the relevant ones to pick, based on the historical data. This feature has saved a considerable amount of time for the auditors.In addition to better efficiency, the AI model also brings value to the consistency of reporting. With around 55 auditors working on individual audits, the need for uniformity and quality control can be fulfilled. The AI model learned from what was previously done and provides suggestions. While more and more data are inputted into the model, it becomes more capable of meeting the quality standards and helping the auditors to do consistent reporting.