Implementing Generative AI for users

Context Like for any software, analytics allow users to understand what's going on and make better informed decisions. That was hardly the case for the users of my customers.

Problem Even though the platform can suggest a wide range of data and tries to transform them into useful information for its users, my customer saw 2 patterns:

1. Users ask +30 times (top 2 most requested feature in 2023) the support team over the last year to export raw data in .csv to manipulate them on Excel

[Side note - beginning]

2 main reasons for this

a. that's because the dashboard didn't provide the info needed (which we solve with the redesign + implemented the export feature)

b. because users want to see correlation with other sources, something we decided to not tackle this way but with third-party integrations for trade-off/benefit purposes)

[Side note - end]

2. Users asked for customer success calls to understand how to interpret data and where to find the most important ones for them. That's a cost that could be optimized and a friction for users.

Solution After some interviews, I delivered the following solutions:

- through a redesign of the information architecture, I classified the information to display the most important ones upfront in Dashboard + 3 timelines selection (custom / last 12 months / year over year)

- redesigned the Dashboard (mainly graph/UI vs rows and columns previously) - implemented an export feature in .csv

- implemented Generative AI, with prompt suggestions at will, so users can ask the questions they were asking to customer success/support.