Google Cloud - Adjudication AI Workshop

Use Case

An HR department representative needs to make a decision on an incident report concerning a specific employee. There is a report from the employee and the employer detailing the incident. Additionally, there are applicable company policies and state employment laws that need to be taken into account. This process is typically arduous and involves the HR representative looking in many places for the data they need to make a decision. 

This application demonstrates how GenAI can automatically bring together various source data based on user input and aggregate it for an employee to review. It also showcases the ability of AI to summarize and make decision recommendations based on a variety of data sources.

Application Description

The Adjudication AI Workshop web application has three distinct components. The first allows data entry from an employee, describing their side of an HR incident. The second allows for input from the employer describing the same incident. Finally, the third component allows and adjudicator to see GenAI summarized inputs from both parties and the applicable company policies and employment laws pulled from GenAI search. It presents all the data in a consumable form to simplify the adjudication process.

Tech Stack

This application is deployed fully in Google Cloud and therefore benefits from the autoscaling, reliability, and security inherent to the cloud. It uses the following Google products:

• Gemini LLM
• Vertex AI API
• Google Kubernetes Engine
• Container Registry
• Firebase