AI and ML at the core of digital transformation in the public sector

Rocío de León, the UK Account Manager at Innova-tsn, looks at AI & ML use in the public sector, which is key to the government’s digital transformation

Article published in Open Access Government on 20 December 2022.

Over the last few years, more and more AI & ML use cases have moved beyond the proof-of-concept stage and are becoming mainstream in the digital transformation of the public sector. Be it making processes more efficient, gaining new insights, or boosting service levels, new use cases are continuously developed and implemented both in the public and private sectors.

A key enabler for governments’ digitalisation

Traditionally the private sector has been a front runner, though governments and public organisations have also picked up pace in the recent past. Despite facing budgetary constraints, the need for improved service to citizens and stakeholders in the public sector is now a part of the roadmap of any digital transformation.


Cloud architectures speeding up use case implementation

One of the key barriers to a broader AI & ML adoption has usually been the lack of the required infrastructure. This is rapidly and drastically changing though. Cloud technologies and the availability of all kinds of tools (e.g., open-source, cloud-native or vendor-specific) are speeding up the migration of on-premises legacy systems and, more importantly, providing the environment required to implement AI & ML.

These platforms are now able to process huge data volumes of structured and unstructured data and provide the tools to efficiently manage the end-to-end cycle of AI & Machine Learning, from data analysis to model development, from fully automated deployments to model monitoring and maintenance.

Analytics and text mining to classify cases in public support services

An example of an innovative use case implemented by Innova-tsn in the public sphere is related to Vincles BCN, the project sponsored by the Barcelona City Council that won the first European edition of Mayors Challenge organised by Bloomberg Philanthropies (a foundation presided over by Michael Bloomberg, former mayor of New York).

As an extension to this initiative, the Barcelona City Council and Innova-tsn developed a solution that allows (more than 400) psychologists in its Social Rights Area to improve the way they assist disadvantaged groups. After gathering problems and needs in an interview, a speech-to-text service feeds an AI algorithm that dynamically classifies the text using NLP/NLU techniques and provides results (in real-time) on the diagnoses and the “most recommended aid roadmap” from the dozen alternatives offered by the different public organisms.

Similar approaches have been applied to other clients for the analysis of phone calls to monitor service quality and adherence to protocols, and identify malpractices or potential fraud cases.

Applying Cognitive Intelligence to move from an attended to a self-service digital framework

The energy sector regulation complexity exposes the different market players to continuous challenges to adapt to, meet and benefit from the established status quo. A dedicated team of experts, to assist those agents with their enquiries and requests to meet the regulation over a complex corpus of more than 400 documents, is a must to ensure regulatory compliance.

Innova-tsn supports one of its clients in an initiative to move to a digitalised environment, unassisted, based on intelligent search and cognitive algorithms to respond to the user’s needs, which customises the user experience and automatically responds to their queries.

“For over 18 years, Innova-tsn has specialised in helping governments, public entities, and corporations of all sizes in applying advanced analytics to make (better) data-driven decisions.”

Application of supervised and non-supervised algorithms to identify potential fraud cases

The Canary Islands Tax Agency improved its fraud detection system, moving from a non-automatic application of business rules to an advanced scoring model and automated alerts. This required the identification of scenarios and fraud patterns (e.g., false billing, company networks, etc.) and the setting of weights and ranges for variables and accelerator scenarios. The process helped improve resource allocation and success rates of fraud investigations.

Similarly, the Spanish Authority for Gambling Supervision (DGOJ) was receiving alerts from operators in the market but wanted to be able to process them more efficiently and set up its own alert system. Different AI & ML approaches were explored considering gamblers and events data, and the best one was selected. Process automation and the possibility to enrich the models with third-party data are now being considered.


Setting the framework for AI & ML implementation in the Spanish Central Bank

AI&ML democratisation and extended application to the entire company’s process lifecycle demand new approaches to model government and collaborative ways of working.

Thanks to its strong analytical capabilities, ranked first in a recent assessment by the Spanish Central Bank, Innova supports this organisation in defining and implementing a new framework that eases collaborative working and traces the whole model cycle, optimising the resources and time to market needed to develop, test, deploy and monitor AI&Machine Learning models. Innova also supports the development of specific analytical solutions following MLOps strategies over the organisation’s platforms.


Innova’s added value

For over 18 years, Innova-tsn has specialised in helping governments, public entities, and corporations of all sizes in applying advanced analytics to make (better) data-driven decisions. AI & ML is just another important building block in building better and faster (even automated) decisions.

Innova’s data scientists combined with the sound organisational and business understanding from its data & business analysts, make it capable of tackling a wide range of analytical challenges. It does so in client assignments but also when addressing government-related challenges (e.g., traffic volume, use of public transport, air quality indices, infrastructure maintenance, crime, etc.) in global Data Science competitions Innova participates in.

However, the successful implementation of AI & ML goes beyond finding the best analytical approach. Innova also supports its clients in designing and implementing the required data and technology architecture (cloud, multi-cloud, etc.) and the most appropriate technology stack and defining the roadmap from the ‘’as is’’ status. Finally, Innova helps organisations in establishing the methodology and framework to manage and exploit AI & ML.