While artificial intelligence is poised to be the next major revolution in the workplace, it remains unclear how such a transformational shift will evolve or affect organisations soon. However, incremental advances in AI are already reshaping operational delivery across our local government contracts
For the Operational Development and Support Team at Serco, AI is not simply a tool for automation and efficiency, but a way to create a more supportive, inclusive, and engaging workplace for our employees. By integrating AI thoughtfully and ethically into our operations, we aim to remove barriers to learning and staff development, increase autonomy, and empower our people.
This means equipping them with the skills and confidence to work alongside AI, enhancing our accessibility offer, and freeing more time to focus on person-centred, human-led training that truly supports our teams.
The training challenge - A challenge we all know well
Training our customer service teams could at times be slow, resource heavy, and inconsistent. Workshops sometimes took weeks to organise and often required staff to spend significant time away from frontline duties. Off the shelf e-learning modules, while convenient, did not always align closely with our specific needs and could feel flat and text heavy, with limited interactivity to sustain attention.
Maintaining consistency across multiple service teams could also be challenging. Updates to policies and procedures did not always reach every team at the same time, meaning not all staff were working with the latest information in real time. This had the potential to create variations in service and affect confidence when handling complex customer queries.
Our workforce reflects a rich diversity of learning preferences and needs, and we're committed to supporting every individual's development journey.
While some employees thrive with conventional formats, others benefit more from visual, interactive, or hands-on approaches. Traditional static resources-such as printed manuals, PDFs, or pre-recorded videos are fixed in format and content, which can make them less engaging or adaptable, especially when trying to tailor learning to different roles, experience levels, or preferred styles.
By embracing more dynamic and flexible learning tools, we ensure that everyone can access content in a way that works best for them.
The costs of delivering and refreshing training content were gradually increasing, both in financial terms and in staff hours. We recognised the need for a solution that could provide timely updates, keep teams engaged, and help ensure every member of staff received the same high quality training experience, across our local government contracts.
Solution
We implemented Synthesia, an AI video generation platform, to create dynamic and interactive training content for customer service representatives. This approach allowed us to develop high-quality training videos without the need for professional filming equipment or presenters.
Additionally, we could quickly customise training materials to address policy updates, procedural changes, and emerging customer service trends. By reducing reliance on text-heavy documents, we increased engagement and knowledge retention among employees.
Implementation
-
Content creation: We produced a series of AI-generated videos covering key topics such as de-escalation techniques, handling complex queries, and using new customer service platforms.
-
Integration with learning platforms: The videos were incorporated into our internal training sites, embedded within instructor-led materials, and included in self-serve modules, allowing on-demand access for staff.
-
Performance measurement: We collected feedback and monitored employee engagement metrics to assess the effectiveness of the AI-generated training.
AI considerations
As with any new technology, the introduction of AI powered training requires careful thought to ensure it enhanced rather than undermined our learning approach. We were clear that our aim was not simply to adopt a new tool, but to maintain the quality, accessibility and clarity that underpin our training approach.
One of the most significant risks was that AI generated content could lack the nuance and authentic insight that comes from the lived experience of our trainers. Without that human perspective, there was a danger that important context might be lost, or messages might feel impersonal. To address this, every video underwent human review and refinement before release, ensuring the tone, content, and delivery met our standards and reflected our organisational standards.
Furthermore, technical precision was equally important. A misaligned voice track, mistranslated caption, or poorly synchronised visual could create poor training experiences, particularly in training on sensitive topics such as customer de-escalation or complex policy changes. We implemented rigorous quality control checks to ensure accuracy, consistency, and alignment between audio, text, and visuals.
We also recognised that over reliance on AI could unintentionally reduce the personal interaction that builds confidence and engagement in training.
To avoid this, we designed a blended approach, combining AI generated modules with live, instructor led sessions where colleagues could ask questions, share experiences, and practice new skills in a supportive setting.
Finally, we were conscious of the risk of ‘content fatigue’. Even the most engaging format can lose impact if it becomes predictable. To keep training fresh, we varied our delivery by integrating interactive proof of learnings, real life scenarios, and discussion prompts that encouraged active participation. This variety ensured that colleagues remained engaged, and that learning translated into improved performance on the frontline.
Results
Impact on future training sessions
The introduction of Synthesia positions us to deliver training that is more flexible, inclusive, and responsive to organisational needs. Content can be updated rapidly to reflect new policies or emerging trends, ensuring staff have timely access to accurate, relevant information. By offering material in multiple formats and languages, we can better support diverse learning preferences and the needs of our cognitively diverse workforce.
Consistent quality across all videos ensures key messages are delivered uniformly, reducing variability in understanding and strengthening service standards. The reduction in production time and costs frees resources for other priorities, enabling us to provide more frequent, targeted training that addresses specific skills or service areas.
The interactive, visually engaging format encourages greater participation than traditional text-based materials, leading to stronger knowledge retention and improved on-the-job performance. As the organisation grows, this AI-enabled approach can scale efficiently, reaching more employees without a proportional increase in cost or production time.
Conclusion
The adoption of Synthesia represents a significant step forward in how we innovate within learning and development. Our priority has been to enhance the employee experience by making training more accessible, engaging, and relevant for every member of our workforce. By adapting to varied learning preferences and presenting information in a clear visual format, we have created a more inclusive and effective way to learn.
This focus on innovation in experience has also delivered important operational gains. Shorter production times, faster updates, and reduced reliance on traditional production methods have led to efficiencies and cost savings, allowing us to invest more in what matters most, our people.
Looking ahead, we see AI not as a replacement for the human element but as a means of continual innovation, keeping training fresh, responsive, and inclusive so that every colleague has the knowledge and confidence to deliver exceptional service.