• Home
  • Projects
  • Reducing inefficiencies: Amey’s next generation AI Modelling transforms Public Sector operations

Reducing inefficiencies: Amey’s next generation AI Modelling transforms Public Sector operations

Abstract representation of digital infrastructure featuring glowing lines, symbolising connectivity.
Speak to an expert about your challenge
Get in touch

Related markets and services

Public sector bodies across the UK face reduced budgets, rising service expectations, legacy systems, staff shortages, and sustainability pressures. Amey, as a leading infrastructure service provider, has identified these challenges and recognised the need for smarter solutions.

At a glance

  • Public sector bodies faced reduced budgets, fragmented data, and resource heavy manual processes for classifying over 50,000 monthly work orders.
  • Amey designed and built an AI driven data classification model using advanced machine learning and Natural Language Processing (NLP).
  • The model automated the manual review of free text work order data, reducing reliance on individual expertise and eliminating bottlenecks.
  • Achieved 90%+ classification accuracy, cut manual workload by 62%, and reduced processing time from two weeks to just 3.5 days.
  • Streamlined operations, improved decision making, and freed teams to focus on higher value strategic work, reinforcing Amey’s commitment to smart technology solutions.

Key metrics

  • 90%

    plus accuracy in classifying work data

  • 62%

    reduction in manual workload

  • 3.5

    day reduction in processing time

We have now developed an automated AI model using Natural Language Processing (NLP) to streamline processes, minimise resource use, and deliver faster, more accurate results. This innovation transforms operations, improves decision making, and reinforces Amey’s commitment to driving efficiency through smart technology solutions.

Our aim was to create a solution that would support better decision making.

What was the challenge?

Across the UK, public sector organisations continue to face mounting pressures. Reduced budgets, rising expectations of service quality, and the need to do more with less impact organisations’ ability to meet their strategic agendas. Many are also grappling with legacy systems, manual data entry, fragmented information sources, and the complex task of maintaining vast estates – challenges further compounded by staff shortages and ongoing pressure to achieve sustainability goals.

Amey, as a leading provider of facilities and asset management services across the public sector, was contracted to support the efficient operation of complex Central Government estates. A key part of this responsibility involved managing repairs and maintenance schedules, ensuring that critical infrastructure remained functional and resilient. However, the process was hindered by the reliance on manual classification of work order data. Maintenance engineers submitted information in free text format, which then had to be categorised, validated, and repurposed by teams. This task was both time consuming and resource intensive, often requiring individuals to manually scan over 50,000 records each month.

The aim was clear: to reduce the burden of manual data management, improve accuracy, and free up skilled staff to focus on delivering operational excellence. Amey needed to find a way to streamline these processes, eliminate bottlenecks, and ensure that data could be processed quickly and reliably to support better decision making and resource allocation.

Our solution balances accuracy, scalability, and efficiency.

How did Amey approach the problem?

To address these challenges, Amey’s data and analytics team designed and delivered an automated AI driven data classification model. Leveraging advanced machine learning (ML) and Natural Language Processing (NLP), the solution was built to process unstructured free text data submitted by maintenance engineers and classify it with high accuracy. This replaced the manual, resource heavy approach with a streamlined, automated process capable of handling large volumes of data at speed.

The expertise within Amey was critical to the project’s success. Data scientists led the design of the algorithm, applying deep knowledge of ML and NLP to ensure the model could ingest and process complex datasets. AI engineers and technical leads contributed to model development, testing, and deployment, while apprentices and graduates brought fresh perspectives and supported experimentation. This multidisciplinary team ensured the solution was both technically robust and strategically aligned with operational needs.

Amey’s approach was innovative and industry leading. Rather than relying on traditional rule based systems, the team explored a range of cutting edge algorithms, including transformer based models and gradient boosting techniques, to achieve optimal performance. The final solution balanced accuracy, scalability, and efficiency, delivering a model capable of classifying work order data with over 90% accuracy. By embedding AI automation into facilities management, Amey not only solved an immediate operational challenge but also set a benchmark for how smart technology can transform public sector service delivery.

We have reinforced our commitment to driving innovation across the public sector.

What was the outcome?

Amey’s AI driven classification model was designed with innovation and operational efficiency at its core, using advanced machine learning techniques to automate complex and repetitive tasks. By removing the burden of manual data processing, the solution enabled individuals to focus on more strategic, high value work.

The impact has been significant. Operational processes in client contractor management have been transformed, with improvements in speed, accuracy, and consistency. The model reduced the strong reliance on individual expertise, delivering accurate and dependable data classification that has cut errors and provided more reliable information for decision making.

Automation has also eliminated bottlenecks, drastically reducing delays and enhancing the flow of processes. Data pipelines are now streamlined, enabling smoother collaboration across departments. Teams can access classified data faster, respond more quickly, and ensure critical information is available at the right time and place.

Time savings have been one of the most tangible benefits. The classification cycle was shortened from two weeks to just 3.5 days, representing a 62% reduction in manual workload. This freed up resources to focus on strategic initiatives and improved overall productivity.

Beyond efficiency, the project has built team expertise. With classification automated, staff are no longer constrained by repetitive tasks and can dedicate more time to creative and strategic endeavours. The collaborative development process, involving apprentices, graduates, data scientists, AI engineers, and technical leads, ensured the solution was both technically robust and strategically aligned.

From day one, the model has delivered measurable impact and established a strong foundation for future automation initiatives.

By embracing intelligent automation, Amey has enhanced productivity, supported growth, and reinforced its commitment to driving innovation and efficiency across the public sector.

Speak to an expert about your challenge.