ai-powered platform for analyzing data and creating reports

ai-powered platform
for analizing data and creating reports

overview

Patterns was accepted into Y Combinator, offering a tool that helps to analyze their company's data and receive custom answers generated by AI. Joining Patterns as a founding Product Designer, I built the product strategy and design system from scratch.

Timeline

Mar 2023 - Apr 2024

My role

Sr. Product Designer

Scope

UX research, wireframes, user flow map, design exploration, visual design, clickable prototype, logo design and branding, design system

Team

1 Product Designer, 1 Product Manager, 1 Front-end, 4 Back-end Developers and QA Engineers

problem

User needs

As a business leader/manager, I need to use a powerful data analysis platform with easy-to-use navigation, so that I can track progress towards goals and identify areas that need improvement.

As a business leader/manager, I need to use a powerful data analysis platform with easy-to-use navigation, so that I can track progress towards goals and identify areas that need improvement.

As a business leader/manager, I need to use a powerful data analysis platform with easy-to-use navigation, so that I can track progress towards goals and identify areas that need improvement.

User problems

Effectively analyzing data is challenging for data analysts and product managers due to complexity of platforms and poor data visualization.

Effectively analyzing data is challenging for data analysts and product managers due to complexity of platforms and poor data visualization.

Effectively analyzing data is challenging for data analysts and product managers due to complexity of platforms and poor data visualization.

solution

Our goal was to create something that would make people's jobs easier and help them get the most out of our tools.

Our goal was to create something that would make people's jobs easier and help them get the most out of our tools.

Our goal was to create something that would make people's jobs easier and help them get the most out of our tools.

design

Initial setup and configuration

User can securely connect to the database and authenticate access permissions. Data is organized and structured in a way that enables efficient querying and retrieval.

Information architecture and visual hierarchy

👨‍💻

Description of this design solution is currently only available on desktop. Switch over for a deeper look!

Scroll the screen below to see design improvements:

Requests

Reports

Home

Type message

Sales by territory chart

View Data

Reports' presentation

Responsive design

design process

Discover

Stakeholders interviews

Market research

Competitive analysis

User interviews

Stakeholders interviews

Market research

Competitive analysis

User interviews

Define

JTBD

Customer journey map

Problem Statement

HMW

JTBD

Customer journey map

Problem Statement

HMW

Design

User flow

Paper sketch

Wireframes

High fidelity design

User flow

Paper sketch

Wireframes

High fidelity design

Deliver

Responsive design

Prototype

Handoff for dev

Gathering feedback

Responsive design

Prototype

Handoff for dev

Gathering feedback

market research

I started research with identifying the gaps and opportunities within the market, which served as a compass guiding our design decisions.

Feature

Notebook-style interface

Drag-and-drop

In-app onboarding tutorial

Interactive dashboards

Natural language queries

AI-assisted visualizations

Automated insights

Predictive analytics

Anomaly detection

AI-assisted code completion

Automated machine learning

AI-generated narratives

Share with team

Creating reports

Version control

Real time collaboration

user interviews

Engaging directly with potential users, we conducted 10+ user interviews and listened to users’ needs, pain points, and desires.

Engaging directly with potential users, we conducted 10+ user interviews and listened to users’ needs, pain points, and desires.

Engaging directly with potential users, we conducted 10+ user interviews and listened to users’ needs, pain points, and desires.

pain points

'how might we' problem statement

Complex setup

Overwhelming initial setup and configuration

HMW we organize the initial setup process to make it less overwhelming for new users?

AI interpretation

Misaligned AI assumptions, lack of transparency, and difficulties interpreting results

HMW provide more relevant and transparant AI insights?

HMW make AI results easier to interpret?

Confusion with indicators

Users are confused by the large number of different indicators like SQL, Data, Tables, and Charts, and struggle to understand how these elements are connected.

HMW simplify and connect indicators like SQL, Data, Tables, and Charts for better understanding?

Lack of clear reporting

No customization, confusing charts, and long duration for data analysis requests

HMW improve report customization and chart clarity?

HMW streamline data analysis to save time?

feature prioritization

I used a Kano model to prioritize features by categorizing them based on their potential impact on customer satisfaction.

I used a Kano model to prioritize features by categorizing them based on their potential impact on customer satisfaction.

I used a Kano model to prioritize features by categorizing them based on their potential impact on customer satisfaction.

design system

I built a design system based on several UX design patterns for creating effective generative AI chatbots. I also utilized resources from Tetrisly UI and Int UI starter kits, both accessible via GitHub repositories.

This snapshot showcases a part of the design system

This is snapshot of a part of the design system

brand identity

I designed a logo, and then developed comprehensive brand guidelines to ensure consistent visual communication across all platforms:

This snapshot showcases a part of the brand guidelines

This is snapshot of a part of the design system

impact

We tracked the success of our design interventions by establishing these key performance indicators:

We tracked the success of our design interventions by establishing these key performance indicators:

We tracked the success of our design interventions by establishing these key performance indicators:

20% conversion rate

20% conversion rate

20% conversion rate

most of users were able to accomplish the intended goals or actions

30% retention rate

30% retention rate

30% retention rate

most of users continued to use the platform over 5 months

Patterns raised $6 million in the latest seed funding round from Y Combinator and Root.vc, affirming investor confidence in our innovative approach and paving the way for further growth and development.

Patterns raised $6 million in the latest seed funding round from Y Combinator and Root.vc, affirming investor confidence in our innovative approach and paving the way for further growth and development.

Patterns raised $6 million in the latest seed funding round from Y Combinator and Root.vc, affirming investor confidence in our innovative approach and paving the way for further growth and development.

reflections

Improving user experience isn't a one-time deal. With technology and user needs always changing, Patterns' platform, built on LLMs, needs to stay flexible. What works today may need to be adjusted tomorrow as user preferences and behaviors shift.

Improving user experience isn't a one-time deal. With technology and user needs always changing, Patterns' platform, built on LLMs, needs to stay flexible. What works today may need to be adjusted tomorrow as user preferences and behaviors shift.

Improving user experience isn't a one-time deal. With technology and user needs always changing, Patterns' platform, built on LLMs, needs to stay flexible. What works today may need to be adjusted tomorrow as user preferences and behaviors shift.

Customers don’t want to wait. Specifically, they don’t want you to waste their time by rewriting prompts dozens of times. If you do make them wait, you risk losing them.

Customers don’t want to wait. Specifically, they don’t want you to waste their time by rewriting prompts dozens of times. If you do make them wait, you risk losing them.

Customers don’t want to wait. Specifically, they don’t want you to waste their time by rewriting prompts dozens of times. If you do make them wait, you risk losing them.

Next project
Next project