Generative AI (Work in Progress)

Goal: Define and Launch GENAI Accelerators

I believe that GenAI is most effective at solving specific, well-defined problems.

Here are the current use cases that I have developed and am exploring for Generative AI:

I developed a detailed pitch and partnered with a cross-functional team for a USCIS demo on how Publicis Sapient can help facilitate a transition from AngularJS to React code in ELIS, using GenAI-driven developer tools
As part of a larger team at Publicis Sapient called GenAI Change Ourselves, I am developing and launching use cases for accelerators in experience design. Currently I am developing two ideas:
- A white label knowledge aggregator that can be applied to various client engagements
- Prototype generators for ideation and design sprints

Some of the challenges that I have observed with GenAI are as follows:

Replicating consistent results
Providing a user-friendly prompt library to ensure the best results
Providing the ability to build on previous results
Setting up back-end data structures to ensure privacy, consistency, reliability and adherence to copyright law and security
Limiting generated designs to a pre-existing design system, without hallucinations

To date I have found that GenAI is better at analysis than repeatable tasks. But we are figuring it out.

Still a work in progress.