AI INTERVIEWING
APP
Target Users
Interviewees
User Size
10M+ Interviews
Platform
Mobile
Company
Avocado Inc.
WHY WE STARTED...
We are 4 enthusiastic people with various high-tech companies background, and would love to try something different. At the beginning, our plan was to create an AI interview tool and sell to different customers. But soon we realized that it doesn’t work, because many of our potential customers can develop by themselves, and unwilling to buy ours.
OPPORTUNITY
After conducting market research, we noticed that the gig economy aligns perfectly with our product. It involves a high volume of interviews, with less personalized or professional questions compared to white-collar roles, and a constant stream of candidates. This model ensures steady revenue, as we charge recruiters based on the number of interviews conducted.
At the time, manufacturing factories faced extremely high recruitment costs. Many built large in-house recruitment departments, renting office spaces for interviews, purchasing testing equipment, and hiring full-time interviewers to conduct repetitive, manual assessments. Others chose to outsource to RPO or BPO companies, which often resulted in complex processes, time-consuming coordination, high costs, and widespread issues like impersonation or plagiarism.
OUR GOAL...
Our goal was to streamline this chaotic process and provide a scalable solution for companies hiring large volumes of short-term or freelance workers—often hundreds or even thousands per day—by automating interviews and improving efficiency through AI.

DESIGN PROCESS

Persona
Use Case
Browser or native app?
We had a general idea that the product should be a mobile video recording interview tool, but we weren’t sure whether it should be a mobile webpage (H5) or a mobile native app.

Considering that native apps need to support both iOS and Android versions, and the complexity and effort involved in maintaining them across different app stores, we decided to start with a webpage instead.
Mini Program
Browser or native app?
Then, we realized that the experience on a mobile web browser was nowhere near comparable to that of a native app. After conducting some research, we discovered MiniPrograms (built-in mini apps within WeChat), which caught our attention.

WeChat is widely known in China, and its MiniProgram supports all devices and operating systems, making it very easy for users to access. It turned out to be the perfect choice for us.
Observing Users
To better understand how users truly feel about our product, I visited the Foxconn recruitment center with our customer success manager, where we immersed ourselves in the entire hiring process from start to finish, observing and experiencing it firsthand.

During the observation, I did find some pain points of users that we will never imagine in the office. For example, a candidate interviewed for the position last month, was hired, but later resigned, and then returned to interview again today, he would fail the screening. However, in reality, he meets the company's hiring criteria.
Information Architecture
Storyboard
Iteration, iteration, iteration...
Take the job video and job description pages as examples, I made a series of iterations for the following reasons:
The header of the page wasted too many spaces without providing useful information, and made the video not immersive enough.
The text information below might be not readable although it already had dark gradient background, especially when video color is light.
Users had to click button below to navigate to job description, the experience is fragmented.
I removed the header and only floated necessary buttons such as back and exit, this makes video areas bigger and more immersive.
I gave the text white background, and use dark text color, so that the contrast pass the Web Content Accessibility Guidelines(WCAG).
I made the white background a scrollable card, users just need to swipe up to view the job description.
The job title and the salary was displayed in the same row, which made the adaptability a big issue. Because the length of the job title and the salary varies a lot.
The differences between the subtitles and descriptions are not apparent, made it pretty difficult for users to read and absorb the content.
Users had to navigate back to the video page to start the interview, which disrupted the flow and made the experience longer and less convenient.
I made the job title and the salary displayed in two different rows, and the title can display in multiple lanes if it is too long. No interference will happen.
I enlarged the subtitle font and applied a gray color to the descriptive text to create a clearer structure and make the content easier for users to read.
I placed the interview button at the bottom of the job details page, allowing users to start the interview immediately after—or even while—reviewing the job information.

FINAL DESIGN

DUO
We later developed an Applicant Tracking System(ATS) to help our enterprise customers track and manage the continuous flow of interviewees. Together, these two products form a complete interview and recruitment experience.

Asynchronous remote interview

Candidates can easily open the MiniProgram by scanning the QR code or tapping a shared WeChat message. Once opened, candidates can commence their interviews. Leveraging AI capabilities, our AI interviewer poses questions tailored to the job requirements and candidate's responses. The interviews are recorded, and upon completion, scores are provided for HR reference.

Manage your applications

Candidates can use their phone number or WeChat account to start and manage their interviews. A list of positions for which the candidate is either in the process of interviewing or has already interviewed will be available.

Job video & details

Prior to the interview, candidates have the option to watch the company's promotional video and scroll down to access additional details about the positions in text format. This information includes details about benefits, salary, locations, job descriptions, requirements, and more.

Abundant types of questions

We facilitate asynchronous interviews using our AI capabilities, where videos are recorded for HR review. Various question types are supported during these interviews.

Mini résumé

We not only support traditional resume documents but also offer support for mini-resume. This means that candidates are not required to upload a meticulously prepared resume; instead, they can simply fill out essential fields before or after the interview.
Shuai Design
2025@Shuai Zhou. All Rights Reserved