Use Case
Customer Service Reimagined
Within the organisation, teams were very busy with Incidents and repetitive tasks, and customer support costs were rising.
​
Time To Assign was hampered by volume;
Time To Respond was impacted by complexity;
Time To Resolve was hindered by inefficiency.
​
Customer Satisfaction and Net Promoter scores were suffering despite creativity and hard work by staff.
​
However, we had a wealth of data at our disposal...
Using Machine Learning techniques (Vectorisation, Semantic Search, Generative AI), we mined Knowledge Articles and comments made in the Client Relationship Manager.
​
This drove improvements in our datasets and customer interactions, and resulted in us delivering practical tools to speed up query resolution.
In parallel, the organisation was starting to focus on its Customer Service metrics, but it had little ability to monitor or predict customer sentiment.
​
We experimented with a variety of Machine Learning models to analyse customer sentiment, and we used experienced staff to train our AI to identify when emails should be escalated.
​
This provided immediate reassurance to managers, and the output was used to analyse and predict the drivers of CSAT and NPS.
But teams had another burden! Some customer Cases were delayed, costly and exposed to risk because staff had to manually read documents, input data and validate information.
We collaborated with operational teams to fuse their specialist knowledge with Machine Learning and rules-based text extraction techniques. This enabled us to implement 'bots' capable of automating critical Case activities.
​
The combination of faster query resolution, sentiment analysis and data entry automation formed a suite of AI tools - using resuable infrastructure - which helped the organisation to implement savings and modernise its customer operations.
​
What's more, we introduced teams to the practical, real-world benefits of these new technologies, and built a network of enthusiastic staff for the next project!
