
GenAI data query platform
COMPANY
ROLE
Expertise
Time & Status
AT&T
UX Designer
UX/UI Design
6 Months, Announced
Project Description
A few months ago, AT&T launched its own generative AI tool called Ask AT&T. This intuitive, conversational platform helps employees across the company by allowing them to interact in plain English (or other languages) to write code and analyze data. It automatically detects fields, joins tables, and generates code to extract insights from the vast data flows managed on the network.
Context:
In June 2023, AT&T launched Ask AT&T, a GenAI platform for employees. It helps summarize phone calls and documents, write code, and extract insights from databases to enhance customer service. The project focuses on querying databases by converting plain English questions into SQL code to uncover data insights.
Team:
2 UX Designers / Product Manager/ IT Project Manager /Developers/ Stakeholders/ Data Analysts
Target audience
Producers
Producers are data owner can create, edit, onboard, and enable tables to consume and enrich data.
Consumers
Consumers are data users who can view table and dataset information and engage in chats.
User statement
Producers
"I want to improve my datasets more easily to get a high reputation score."
Consumers
"Even though I'm not the owner of this data, I'd like to view and chat with the bot to gain some insights."
Problem
Users often feel frustrated when trying to translate human language into query code and technical terms, especially if they are not data experts. This can lead to inefficiencies and ineffective workflows.
Goal
The tool empowers AT&T employees who aren't data experts or programmers to write code and algorithms, extracting actionable intelligence from complex data sets. This will enhance the efficiency of customer service agents, allowing them to address customer questions more effectively and swiftly.
​
Constraints
-
Ensure that this design aligns with AT&T's existing Ask Data design system.
-
This project included numerous technical terms related to engineering and coding. To bridge this gap, we worked closely with UI developers who translated these terms into effective UX design, ensuring a seamless and user-friendly interface.
Ideation
​Imagine a Data Processing Inventory experience that supports member self-service and delivers trustworthy content, intuitive navigation, and improved usability.
Solutions
​Imagine a Data Processing Inventory experience that supports member self-service and delivers trustworthy content, intuitive navigation, and improved usability.
Feature 1: Improve Datasets Flow
We created an enriched dataset/table flow to achieve a high reputation score and improve accuracy.

Feature 2: Chat Flow
Interact with the AI bot to help users get answers using SQL queries, code explanations, and charts/graphs from plain language questions. Chat w/ Data is visible to all users.

Version 1

Version 2
Version 3


Feature 3: Create new dataset (multiple tables)/ new single table

Create a new datasets

Create a new single table
Feature 4: Connect Table/ update table(s) in dataset

Connect Table

Update Table(s) in Dataset
Version 1
Version 2
Version 3



Feature 1: Improve Datasets Flow
We created an enriched dataset/table flow to achieve a high reputation score and improve accuracy.
No advanced search
Iteration
We hold demo meetings with developers and stakeholders every two days. These mock-up sessions foster close collaboration among all teams. Our UX team actively seeks feedback from stakeholders, which helps us adjust the design accordingly.
​
Learning
AT&T has been a pioneer in AI for decades, and that commitment continues today. By empowering employees with innovative tools, we enhance their creativity and effectiveness.
Through this project, I learned the importance of prioritizing work under tight deadlines. This experience underscored the critical role of communication in cross-functional teams, ensuring everyone is aligned with project goals. I also recognized the value of integrating diverse perspectives, which enriched our discussions and led to more effective solutions. Overall, this project deepened my appreciation for teamwork and collaboration in a rapidly evolving lanscape.