The food pantry at Georgia Tech, Klemis Kitchen, is the main food assistance program on campus which provides both undergraduate and graduate students who could not afford balanced meals with free food items. However, Klemis Kitchen has long been lacking an online presence and relying on emails for communicating with its users. We designed an app to provide students with a digital experience to enroll in the service, check the availability of the food items and engage in the Klemis Kitchen community as a whole.

2021.08 - 2021.12
UX designer &
UX Researcher    
Figma, Qualtrics, Miro
Team Memebers
Greg Parker, Jessie Chiu, Srijan Jhanwar


The admin's "magic email" serves as the only communication tool with Klemis users.

However, based on our surveys and interviews, users found the information in these emails insufficient and confusing, especially in terms of the availability of food and the enrollment process.



First-time users are able to browse what services Klemis Kitchen offers, submit the enrollment form, pick their dietary preferences all at one place.

Users could easily see how much food is left in the pantry from the color-coded labels "Plenty""Limited" and "Out of Stock" on the main page. They could also check the nutrition facts and whether the items meet their dietary restrictions in detail.

To support long-term food security and healthy eating, we provide users with various recipes based on the available food in the pantry.

Interactions between Klemis users are supported by the "Community" tab where users chat in threads, making admin announcements effortless and creating a sense of community to empower students suffering from food insecurity.

Our Process



We sent out surveys to all 40 current Klemis Kitchen users and the responses show that most of the current users found out the food pantry by word of mouth from the school staff or friends. Along with the fact that Klemis Kitchen does not have a website of its own and its social media account has been inactive since 2015, we found the evidence a strong indicator that Klemis Kitchen significantly lacks an online presence. Therefore to put its information online and to digitalize the enrollment process caught our eye as an implication for our design.


We conducted 4 interviews with both the service providers and the student users and we used affinity diagramming to analyze the data. The key findings are as follows:

  • Users of Klemis Kitchen found the onboarding process unclear and fragmented.
  • There is a need for a real-time view of available foods, clear nutrition labels, and expiration information.
  • Email is the only source of information on food availability, which users find inadequate in supporting timely notifications.
  • Users feel indebted to Klemis Kitchen and don't want to be a burden to the staff and other users.
  • The social stigma around food assistance discouraged users from seeking help early.
  • Klemis Kitchen has infrastructural and logistical constraints within which our design will have to operate.

Affinity Notes from Interviews with Student Users

Affinity Notes from Interviews with Service Providers


We went to the food pantry to do an on-site observation and to talk to the Klemis Kitchen volunteers. The space of the pantry is very limited with one fridge, one freezer and four shelves, two volunteers were working in their shifts and the whole operation does not involve any technology. The lack of space and man power uncovers the underlying reason for why there is not a real-time inventory: There is no space to give a clear view of all food and limited volunteers do not have time for inventory management. The key findings are as follows:

      Abandoned Checkout Kiosk

We asked the volunteers and they said that this checkout/POS system has never been used even though they have the machine. So they have never kept track of how much food students took.

      Unlabeled Food Items

There are many food items manually packaged without any labels, such as rice. Volunteers portioned the food to accommodate more students but this prevents users from knowing what's inside.

      Paper Recipes

There is a brochure hung on the shelf which provides simple recipes containing the ingredients from the pantry, helping users learn how to cook with the food they get.

Hierarchical Task Analysis

Based on the surveys, interviews and on-site observation, we narrowed down our problem scope to focus on these two processes with the most issues revealed:

  • Getting access to the food pantry/onboarding
  • Using the food pantry onsite

We conducted 2 hierarchical task analyses to gain a better understanding of the tasks users need to complete to achieve these two goals.

We found that, for getting access to Klemis Kitchen, where we could optimize the process is to combine the sub-tasks of seeking help, contacting the organizer and getting access to the pantry into one integrated onboarding process on our app. So that users could complete them through a couple of taps within five minutes instead of looking for various sources of information, emailing the organizer back and forth, and experiencing the frustration of waiting for a week or so.

For using the food pantry, we found that the main task that our system need to support is displaying the inventory and the details of the food items for users to find their desired food more efficiently.


Ideation & Sketches

Based on our research findings, we did a brainstorming session and sketched out the top 10 ideas to quickly visualize the concepts. After discussing their pros and cons in terms of meeting users' needs and functioning under the logistical constraints, we decided to incorporate 3 of the 10 design ideas to move forward with, which are:

      Group Track

Group Track is a crowdsourcing feature that allows users of Klemis Kitchen to update the inventory of the food pantry themselves.

      Cooking 101

Cooking 101 is a feature that suggests recipes which can be prepared with the items that users are picking up.  

      Take Your Pick

Take Your Pick feature allows users to select their dietary preferences and food preferences as a part of the onboarding process.

Other 7 design ideas:

High-Fidelity Prototypes

      User Testing


I developed the testing script and led the three user testing sessions with our participants recruited from Georgia Tech students. For each testing session, the participant is given 5 task scenarios and they think aloud while completing these tasks. The tasks are:

After each task, there is a structured interview where we asked questions about our users' thoughts and feelings on the task they accomplished. The user testing sessions were conducted on Teams as an accommodation to the pandemic situation.


At the end of the interview, we gave users 4 Likert scale questions to understand their attitudes towards the app and we found that all users at least somewhat agreed that they got all the information they need to enroll in the service, felt that the onboard process is transparent and were satisfied with their onboarding experience. In addition, all participants expressed that they did not want external help while navigating these features and this potentially indicates a good usability of our design.

Integrating the data from all testing sessions, we have some major findings that give design implications for future iterations:

  • The onboarding process is overall satisfactory but requires more detailed informationto provide context for new or potential users of Klemis Kitchen.
  • Unconventional user flows and poor affordance causes unnecessary confusion and frustrations.
  • Participants found it easy to discover the availability of food but faced difficulties completing the subsequent actions.

Lessons Learnt

      Embracing Constraints

At the research stage, we interviewed not only the users but also the service providers which include the Klemis Kitchen staff and volunteers to meet the needs of both stakeholders. According to these interviews, we found that there are a lot of physical and technological constraints in the food pantry which we need to take into consideration in our design, such as the limited space and budgets.

So although we had many wild ideas, I had to pick the ones that best fit their real circumstances. For example, I advocated to go with the Group Track idea instead of the autonomous inventory system idea because I understood that the current budget would't allow the food pantry to make the system innovation expected in this design and thus to let student users collaboratively update the inventory would be the more cost-saving and feasible solution. This approach turns out to generate extra benefits for the users because the crowdsourcing format also enables them to give back to the pantry and feel a sense of community, which cannot be obtained from the autonomous system.

From this experience, I gained a deep understanding of how to navigate the constraints in UX design and I actually appreciated the constraints as they pushed me to jump out of the box and solve the problem more creatively.