The authors want to thank Liesbeth Mulder, Aon Connect global reporting lead, Rodrigo Bernardes, Aon Connect global reporting analyst, and Raja Pratap Kondamari, product manager for OCI Language, for their contributions.
Figure 1: Aon office building
Aon plc (NYSE:AON) is a leading global professional services firm providing a broad range of risk, health, and wealth solutions. Their 50,000 colleagues in more than 120 countries and sovereignties empower results for clients by using proprietary data and analytics to deliver insights that reduce volatility and improve performance.
The company has increased focus on customer relationship management through digital transformation. Aon was looking for more insights related to their client feedback surveys and specifically analyzed sentiment related to their responses to gain broad insight into trends of how their products and services are received and where the areas for improvement exist.
Goals for cloud migration
As part of their digital transformation, Aon is modernizing and centralizing their customer service. The company had accumulated a large volume of data but had no practical method to rapidly and reliably extract useful customer feedback to optimize its business and offer new products. To address this challenge, Aon has chosen Oracle Cloud Infrastructure (OCI), taking advantage of OCI Artificial Intelligence (AI) and other services for large-scale data processing. The company had the following critical goals aligned to this initiative:
- Aon receives thousands of customer feedback forms and reviews every month and accumulated large volumes of such data over the years. They wanted to use that data to gain insight into how to improve their products and services. This process was challenging because the data was unstructured in the form of text-based reviews, making it challenging to consume and analyze quickly and consistently in large volumes.
- Aon sought to better understand the sentiment of their customers and quickly identify the keywords that signaled different sentiments and the entities, such as products, services, and organizations, that garnered positive or negative feedback.
Suite of Oracle products used
The company used the following OCI services, software development kits (SDKs), and frameworks:
- OCI Language: Part of the OCI AI offering, OCI Language makes it possible to perform sophisticated text analysis at scale. With pretrained built-in models, developers don’t need machine learning (ML) expertise to build sentiment analysis, key phrase extraction, text classification, named entity recognition, and more language AI capabilities into their applications.
- OCI Autonomous Data Warehouse: A fully automated database service that makes it easy for all organizations to develop and deploy application workloads regardless of complexity, scale, or criticality. Autonomous Database’s converged engine supports diverse data types, simplifying application development and deployment from modeling and coding to extract, transform, load (ETL), database optimization, and data analysis.
- OCI Analytics Cloud (OAC): The Oracle Analytics platform is a cloud native service that provides the capabilities required to address the entire analytics process including data ingestion and modeling, data preparation and enrichment, and visualization and collaboration, without compromising security and governance.
Customer’s solution on OCI
The total implementation time from proof of concept (POC) to production workloads was approximately 25 days.
The architecture pattern includes several steps for managing and interpreting data: Discover, ingest, transform, curate, analyze, learn, predict, measure, and act. The overall process flow includes the following steps:
- Data is replicated from Sales Cloud using Oracle Analytics Cloud (OAC) and stored in Oracle Autonomous Data Warehouse.
- Data is preprocessed using the inbuilt SQL and PL/SQL capabilities of Autonomous Data Warehouse.
- The OCI Language service API is called from Autonomous Data Warehouse through the PL/SQL service and performs sentiment analysis.
- The result of the analysis from OCI Language is stored in tables in Autonomous Data Warehouse and prepared for visualization in views.
- The end users access these views through a data model in OAC and use them to develop visualizations that deliver actionable insights.
Figure 2: Architecture diagram
With data driven insights, Aon can now easily identify the top priorities of their customers and monitor customer sentiment. With the data pipeline built by Oracle Consulting, Aon can consume the sentiment analysis functionality in OCI Language, convert unstructured textual content to structured content, and present it as interactive dashboards, so that colleagues can quickly access insights and drive action.
By utilizing OCI AI, OCI Autonomous Data Warehouse, and OAC, Aon can automate the processing of customer feedback, improve business efficiency, and increase customer satisfaction.
Figure 3: Sample Dashboard of Analytics
Next, Aon intends to extend the use of OCI’s AI capabilities to more use cases for sentiment analysis. The company also plans on exploring the potential of other Autonomous Data Warehouse services for analytics, notably Oracle Machine Learning for predictive analytics.
For more information on Aon and Oracle Cloud Infrastructure, see the following resources: