P R O J E C T
GenAI Customer Review Sentiment
A generative AI data project to extract useful sentiment and classification from raw customer reviews
B A C K G R O U N D
As a company we were ingesting reviews from both Trustpilot and Google Reviews where we get the title, rating and content of the review.
In monthly updates we had somebody manually sifting through these reviews and extracting metrics by eye, establishing things that customers highlight as particular concerns and key celebrations that our users express.
We wanted to reduce wasting specialist time on menial tasks so decided to offload this work on an generative intelligence agent then we can automate the monthly customer insights dashboard.
T E C H N O L O G I E S
C H A L L E N G E S
Company buy in was more difficult than expected for this project. Given the industry excitement surrounding generative AI at the moment, the business was worried about becoming bought into one solution too early and needing to replace it down the line.
This was shortly after another set of tooling increasing their prices so there was concern that cost of these AI tools will increase with time.
There was also the technical unknowns in the team, many had experimented with AI in personal projects but not on the scale that was required in this instance.
P R O J E C T
O U T C O M E
What would have previously taken a colleague a few hours to collate has now been complete in just a few minutes.
The team's excitement prompted a number of ideas for GenAI innovations that were added to the project roadmap.
We managed our resources carefully and resulted in a daily running cost of less than 1¢.
P E R S O N A L
G R O W T H
From this project one of the big take-aways is to think outside the box when considering data stakeholders, teams often get in a rut thinking they report to business analysts however there are opportunities to streamline work for everone around the business who needs to interact with data at any level.
We planned this project to be a small dip into a big subject and to gain stakeholder trust in the product. This paid off massively with a burst of GenAI enthusiasm in the company stretching beyond the data team. We will be applying this approach in more areas to encourage similar enthusiasm about changes in the industry.
