startupinternz.com

Job matching program for startups and students

Overview

Date: 2016 ~ 2019
Main Purpose: job matching program web platform
My Contribution: Project Management UX Design Web Design Web Development

Description

As a product manager and software engineer, I built and improved Startupinternz from scratch. Korean startups often find it difficult to find suitable candidates. Job applicants, on the other hand, find it difficult to obtain useful information about startups. Startupinternz was created to solve this lack of information on both sides, as an educational program that could help college students work with startups that best matched their abilities and character.

Also, I implemented a core algorithm of the matching service for Startupinternz users. First, I researched to understand which factor most influences users' decisions when they seek jobs, and it is found out that 'cultural fit' is the underlying interest users most value. The work process from research, programming, and multiple iterations has ultimately improved the general user experience of Startupinternz as the applicants who used this matching algorithm showed higher retention rates than students who had not.

Based on the lean startup methodology, I made a one-page website -- a minimum viable product -- to collect additional user feedback. I developed the website further, into a personalized web application, then into an admin application which helped the platform become scalable through automating its administrative functions. As a result, Startupinternz has expanded over the years into a popular platform for both startups and potential employees, hosting 43 offline workshops in four years and helping 486 (and counting) applicants successfully begin their careers at various startups. The project taught me the entire iterative process, from ideation to implementation, then to deployment. It was also an opportunity for me to develop a scalable platform by applying lean startup methodology to an actual product.

2019, Improved Bouncing Rate

Overview

Date: 2019
Main Purpose: job matching program web platform
Project Size: Multiple Producers, 2 Web Programmers and 1 Web Designer, 6 weeks
My Contribution: UX Design Web Development
Stacks: HTML SCSS JavaScript jQuery Ruby on Rails PostgreSQL Heroku AWS S3

part of landing page

Description

It is my last project at Oeclab. I fixed the bouncing rate issue of startupintern.com by making modifications to rearrange information and apply new designs.

Takeaway

  • By implementing a new design and IA, we were able to get a better result in the bouncing rate and number of users. It has been confirmed again that viewing from the user's perspective has strong consequences.

2018, Upgraded Business Pages

Overview

Date: 2018
Main Purpose: job matching program web platform
Project Size: Multiple Producers, 1 Web Programmer and 1 Web Designer, 8 weeks
My Contribution: Web Development
Stacks: HTML SCSS Bootstrap JavaScript jQuery Ruby on Rails PostgreSQL Heroku AWS S3

part of landing page

Description

For specialization, the Startupinternz team designed the website and features, and I participated as a developer.

Takeaway

  • For specialization, I performed one role, not a multi-role. I managed many communication nodes, working with several Startupinternz team members and a designer. I focused on documentation to reduce miss communication.

2017, Multi-page Web Application

Overview

Date: 2017
Main Purpose: job matching program web platform
Project Size: Solo Project, 10 weeks
My Contribution: Project Management UX Design Web Design Web Development
Stacks: HTML SCSS Bootstrap JavaScript Ruby on Rails PostgreSQL Heroku

part of landing page: page provides information about the program

apply page: users can download and submit application form

Description

Startupcampus has been renamed Startupinternz. I have implemented a multi-page website that offers more functionality by reflecting customer feedback and internal feedback. The first landing page showed the information the audience needed, and the other pages provided more detailed information. It provided the functions needed to operate a ‘Startupinternz’ training program, including applying, submitting assignments, and reading reference materials. Also, the administration has been able to manage schedule programs, references, and Q & As.

Takeaway

  • The 20s were mainly targeted users. Because they prefer mobile devices, I decided to build a mobile-first web site. According to the analyzed data of startupinternz.com, over 80% of first-comers used mobile browsers.
  • I learned a lot about Information Architecture. Because it was a new type of program that did not previously exist, there was a lot of newly generated information to provide for the users. I tried to organize and convey the information easily from the perspective of the user.

2016, One page MVP (Minimum Viable Product)

Overview

Date: 2016
Main Purpose: job matching program web brochure
Project Size: 1 Web Programmer and 1 Web Designer, 4 weeks
My Contribution: Project Management UX Design Web Development
Stacks: HTML SCSS Bootstrap JavaScript Ruby on Rails Heroku

part of website

Description

A website for applicants of the ‘Startupcampus’ program which was a new business of Oeclab corp. I worked with Ruby on Rails away from LAMP (Linux, Apache, MySQL, PHP) I skilled. I started with a simple one-page website, as Lean Startup theory. I minimized the development resources by leveraging Dropbox's file request feature.

Takeaway

  • I used Ruby on Rails for the first time in a product. It was possible to develop faster than a LAMP (Linux, Apache, MySQL, PHP) used in existing work. As an alternative to LAMP, I considered Laravel, Django, and Ruby on Rails, and chose Ruby on Rails to better suit the company's situation. I was able to develop a web application quickly by using Ruby on Rails due to its diverse gem ecosystem, and it was easy to find documentation. Also, because I had to create and manage multiple websites on my own, the convention over configuration (CoC), which forces a lot of work in the framework, was a big advantage.
  • I took advantage of appropriate alternatives. There are a lot of things to consider to create a function that receives files from a user. To prevent hacking, you need to restrict the types of files uploaded, limit the file size, and manage the permissions of the uploaded folders. The file received from the user should be made available to the administrator, and the function to be downloaded and used should be created. Much of this was easily solved through Dropbox's file request feature.
  • Alternative features couldn't satisfy everything. I tried to provide an application file using Dropbox and Google Drive, but I was not able to satisfy with UX. In the end, the application file was provided directly from the server.