Finance professional services institute

Finance professional services institute "goes digital"

An international finance professional services institute wanted to undertake a review, classification and merging of its membership and image data to help the organisation ‘go digital’ Aim was engaged to help




An international finance professional services institute had grown organically over the past 10 years delivering qualifications, training and governance certifications in a range of financial services and pension subject areas. The qualifications and certificates granted have an international reputation and draw candidates from across the world. 


But that organic growth had never led to the creation of a serious data and integration strategy, such that by 2019, the administration teams were accessing tens of thousands of membership records and qualifications data via 5 different systems. This was not only becoming administratively unmanageable, but errors were being made, records were being lost or duplicated and the institute’s auditors had made an observation about the implications for data protection compliance. Data protection was important as the institute held a considerable amount of personal identifiable and sensitive information (PII), which included text-based information such as credit card numbers, as well as images and audio files. Images related to personal identification documents such as passports, ID cards, driving licences, visas etc – some of which were digitised whilst others were still held as hard copy.  Audio files related to phone calls recorded with members.


The institute decided to improve all these areas as part of a ‘going digital’ initiative and asked Aim using its dataBelt® data governance platform to assist with this transformation.




Aim data analysts created a data strategy that would drive forward a ‘go digital’ project for the institute. A key work stream of this project was to resolve the membership data and data protection aspects that needed priority attention. The institute had already recognised the need to bring the 5 membership databases together and had settled on convergence to a single system, selected as part of a separate product review process.

Aim implemented its AI-powered data governance platform dataBelt® to deliver the following:

·        Discover and index all the data held by the institute – structured and unstructured;

·        Digitise the hardcopy data and images;

·        Classify all the data in any format and location, which for databases involved drilling down to the table and column level to understand type and category;

·        Build a data inventory to support the data strategy;

·        Merge the 5 databases into a single dataset for migration to the new single membership system, cleansing, de-duping and ensuring there were no orphan records; and

·        Provide the understanding and ability to comply with data protection legislation (eg GDPR) and ensuring there was no data hoarding, ie data that could no longer be justified as being stored.


Aim used dataBelt® Keras Tensorflow machine learning and optical character reading (OCR) tools to discover images and audio files held by the institute and to classify them. Classification was made in terms of textual information where available and to recognise images of different types. A number of models were used to train dataBelt® machine learning (some already available in its model library) so that image and audio identification could be made quickly and accurately.




The analysis and outputs that Aim produced via dataBelt® delivered on all aspects of the data transformation assignment and created excellent results that would have taken the institute months, if not years to complete via other means. It ensured that the institute worked to a data strategy, single, clean membership dataset, and was data protection compliant – closing the audit observation.


In particular, the ability to use new AI technology to understand images and audio files was seen as a big win and the institute has approached Aim for a phase 2 of the assignment to help with video footage of training sessions and examination vivas.