Algorithm Platform

2021 Summer, Enterprise UX, Shanghai

Duration: 3 Months

My Role: UX Design & UI Design

Collaborated with Qinhao Zhang, Xin Yang

Brief Introduction

The Algorithm Platform is a website for showcasing Ping An Technology's 100+ AI algorithm products. I designed the website to help users quickly find a solution to their business needs.

Background

Ping An Insurance (Group) Company of China is considered the largest insurance company in China. Ping An Technology, a wholly-owned subsidiary of Ping An Group, focuses on providing intelligent technology services to both the Group's internal organization users and external enterprise users.

The service scope covers over 95% of Ping An Group's companies and supports 80% of business systems in production. This company's services span finance, medical, automotive, and environmental protection. The core technologies include biometric identification and face recognition, blockchain technology, cloud services, etc.

Problems

1. Ping An Technology has many algorithm products but lacks an integrated platform for collecting, organizing, and displaying algorithm product services. As a result, obtaining algorithm product resources is challenging, and the process of using them is cumbersome.

2. As a subsidiary of Ping An Group, Ping An Technology has a complex group of users, including internal employees and external users. Different user groups have very different goals and processes for using algorithm products.

3. The number of algorithm products owned by Ping An Technology increases rapidly, which causes difficulty in organizing.

What we did

1. Build an easy-to-use web platform to access the 100+ algorithm products.

2. Design different usage processes for different user groups to improve the efficiency of the service.

3. Improve the usability and scalability of the algorithm search system.

Our solution

Meet Ping An Algorithm Platform.

Ping An Algorithm Platform is a comprehensive algorithm platform for external users, internal users, internal developers, and internal managers, providing all user groups with Ping An Technology's three core services through a web-based format.

Background Research

Based on company management's vision for the algorithm platform, the algorithm platform will provide the following three core services to users.

AI incubators

Provide integrated enterprise management services (including human resource management, financial resource management, and information resource management) for algorithm developers of small and medium-sized enterprises.

AI services

AI algorithms include machine vision, text recognition, data analysis, risk analysis, etc.

Industry-specific solutions

Integrated algorithmic services for specific topics, e.g., education, retirement, investment, urban construction, etc.

Target Users Research

1. External users

Small to medium-sized technology companies that do not yet have sufficient technology and resources to develop their algorithm products need to purchase them.

2.  Internal users (ordinary employees)

Ordinary employees within Ping An Group need to use the algorithm products within their access level according to their tasks.

3. Internal users (management employees)

Management employees within Ping An Group have the right to manage and assign algorithm products and can open algorithmic product access to their subordinate employees.

Journey Map

We researched 108 different users of the three categories above by observing and asking them about their attitudes towards the existing query process for algorithmic products, then mapped the user journeys of typical users.

Market Research

I researched the algorithm platform websites of Chinese large-scale Internet companies, clearly understanding the architecture of competitor websites.

I also did in-depth research on the design of the most influential websites to target the website's visual style and effect.

Information Architecture

Combining the uniqueness of the company's situation, the problems to be solved, and the market research, the design team determined the website's structure: homepage, catalog page, details page, console page, and dashboard page.

The dashboard page differs from other websites based on the market research results. It is mainly open to internal users to upload, manage, use and observe algorithm products.

Design iterations

The Algorithm Platform website is a new user interface and user experience design project from scratch; each page has gone through 4 to 5 iterations; this picture shows part of the iteration process.

Main Pages: Homepage

Design goals: Help users understand the scope of services offered by the algorithmic platform.

Main Pages: Catalog Page

Design goals: Help users search for the algorithm products as efficiently as possible.

Pic11_ AP-catalogue.png

Main Pages: Details Page

Design goals: Help users to get comprehensive information about the algorithm service and simplify the purchase process.

Pic13_AP-details.png

Main Pages: Console Page

Design Goal: Help users to visualize the usage and purchase of existing algorithmic services.

Pic15_ AP-console.png

Main Pages: Dashboard Page

Design goal: Help users get the algorithm service and the revenue information at a glance.

Pic17_ AP-dashboard.png

Workflow

Design System

I built the whole design system with another team member Qinhao Zhang, using Cinema 4D, Rhinoceros, and Keyshot to finish the 3D modeling design and rendering.

What's Next?

The algorithm platform is a large-scale project. I am fortunate to participate in and be responsible for the user experience design framework and essential page user interface design.

I used to think that research only existed before the design. However, in this project, every iteration of interface design is inseparable from research, which helped me clarify current design problems and provide improvement ideas.