Few types of data are more valuable to a business than contact center data. Contact centers offer a direct touchpoint with customers, giving organizations insight into their needs, and an opportunity to earn their loyalty.
When Deloitte Canada was tasked with transforming a client’s call center , the goal was to extract and translate highly custom insights into a data-driven roadmap.. However, as they delved into the project, the sheer volume of customer interactions in the form of call center recordings and transcripts proved overwhelming. Analyzing and extracting insights from this raw data using manual processes proved to be time-consuming and costly, so Deloitte realized it needed a more efficient system to explore, analyze, and perform topic modeling.
A streamlined approach to search, cluster, and easily uncover trends
Deloitte collaborated with HumanFirst, which provides advanced natural language understanding (NLU) capabilities designed to extract insights from within raw, unstructured data in just a few clicks. Leveraging the platform allowed the Deloitte team to automate the processes of clustering similar transcripts and topic modeling.
Using HumanFirst’s NLU Design and Google Cloud’s robust cloud infrastructure, Deloitte was able to build a strong foundation for storing, managing, and accessing the client’s call center data. This direct integration between HumanFirst and Google Cloud ensured the scalability of the dataset without compromising Deloitte’s data modeling accuracy or performance.
HumanFirst achieved this feat by utilizing Google’s Kubernetes Engine, which ensured a foundation of reliability, hands-off operation, and secure-by-default design. The architecture enabled dynamic scalability of the HumanFirst product infrastructure. Deloitte benefited from the collaborative features embedded in HumanFirst that streamline data exploration and insights in addition to the effortless integrations with Google’s suite of applications, which can readily adapt to demand spikes. This enhancement significantly improved the team’s end-to-end workflow and process efficiency.
Deloitte’s customer mandate to execute a digital transformation was streamlined as HumanFirst enabled fine-tuned prompts to extract call drivers from the large conversational dataset. Additionally, it unlocked data insights that informed the structured ontology by applying clustering and semantic search techniques effortlessly within the platform. By utilizing Large Language Model (LLM) data transformation pipelines, they were able to elevate the depth of analysis within HumanFirst. These efforts culminated in a wealth of insights into call drivers and resolution journeys. The outcome was a transformation strategy that zeroed in with precision, all thanks to the enriched understanding gained through these processes.
“This flexibility proved crucial in handling our client’s dataset and is accommodating for future or similar projects,” Anand Nimkar, Chief Architect of Generative AI at Deloitte Canada, said. “As our client continues their journey with Google’s CCAI technology, there will be opportunities to further leverage HumanFirst and its integrations with Google’s digital ecosystem.”
Enhanced efficiency and actionable insights at a fraction of the cost
After implementing HumanFirst, the time required for topic modeling data saw a staggering 90% reduction, which made analyzing large volumes of call center transcripts efficient with fewer errors and revisions throughout the review cycle.
“The time spent on manual analysis was drastically reduced, enabling the team to focus on higher-value tasks,”Nimkar said.
With HumanFirst’s built-in AI features that leverage Google Dialogflow’s NLU and Vertex AI’s LLM capabilities, users are able to efficiently explore, transform and understand large amounts of unstructured data.
Deloitte leveraged HumanFirst’s topic modeling tool which accelerated its approach to understanding call drivers in their client’s contact centers. Going forward, Deloitte will continue to introduce HumanFirst’s tool and Google’s suite of technology to our other clients who need similar analysis across unstructured data. Deloitte was able to surface insights and trends within the client’s customer interactions and identified common pain points and call drivers. Additionally, Deloitte was also able to spot areas to improve agent training and processes, as well as AI automation opportunities.
The implementation also led to a remarkable 80% reduction in costs associated with the standard data analysis processes.
“It previously took a team of four people to perform topic modeling over five days using our traditional analysis methods; whereas with HumanFirst, we only needed two people and performed the same amount of work in one day,” Nimkar said. “This allowed our engagement team to deliver a high-quality solution to their client within budget constraints.”
Google Cloud, Deloitte, and HumanFirst accelerate contact center transformations together
Deloitte’s collaboration with HumanFirst and Google Cloud was able to drive positive changes in the customer’s call center operations. By automating and streamlining the exploration of valuable insights on Google Cloud architecture, Deloitte not only significantly enhanced efficiency, but also transformed its data analysis process.
Learn more about Google Cloud’s open and innovative generative AI partner ecosystem and about Deloitte and HumanFirst on Google Cloud.