Allervys
An employee-facing food-safety co-pilot that uses computer vision to flag contamination risks in real time. It gives crews instant, step-by-step fixes and managers a verifiable audit trail—protecting diners and the brand in one seamless workflow.
Project at a glance
Team
Solo Product Designer
Time Period
Feb 2025 - April 2025
Tools
Figma
FigJam
Notion
Chat GPT
Hunter.io
Project Type
Final project for an Innovation, Design, & Entrepreneurship course
Hypothetical product
Food allergies suck.
On a late-night Gelato run, a staff member admitted they couldn't guarantee a nut-free scoop for my friend Jack. You see, Jack has a severe nut allergy, therefore he didn’t want to gamble with a possible trip to the ER, so he walked out empty handed, and Figo lost a potential customer.
This is the only photo I have of us at Figo, Jack is the blonde in the back.
And as it turns out , over 50% of allergic reactions strike even after guests warn staff.
The root cause is bigger than just labels, it’s real time employee lapses.
Initial Research
What I thought the problem was
At this point in the project, I knew I wanted a narrow focus on stopping or resolving allergen cross-contact in bar-style kitchens.
[insert graphic]
1 in 3 fast-casual kitchens still cross-contaminate in plain sight
[insert graphic]
One slip can cost a fast-casual $2.1 million, before brand damage
From a user and business context, this made me realize it would be a waste to be focus on simply allergen cross-contact, so I broadened my scope to cross-contamination and food safety in general. The value prop? Building customer confidence and slashing legal risks.
Research Insights
Getting started with validaiton & research
It was time to start getting insights from professionals & buyers in the field.
After I got some roles, I went to LinkedIn to do cold outreach and scouted emails using Hunter.io, in total I contacted over 40 people.
While I waited I created some user personas to help me get an glimpse of what some user needs may be.
Personas
End User
Tobey
General Employee at a Cafe
About
Third-year design student who works the evening barista shift. Tech-savvy, quick to learn new routines, and takes quiet pride in crafting consistent drinks for regulars.
Responsibilities
Making and preparing specialty drinks, coffee, as well as mini sandwiches and sweet treats. In charge of ensuring quality and safe handling of food.
Frustrations
Sometimes slip up with store tasks like washing utensils, changing gloves. Additionally, theres knowledge gaps like awareness of allergens.
Personas
Buyer
Celia
Cafe Owner
About
Thirty-something entrepreneur who turned her love of third-wave coffee into two bustling cafés. Passionate about local art, mentors junior staff, and starts every morning dialing in the espresso herself.
Responsibilities
Ensures the business stays profitable, takes care of legal requirements and issues. Makes sure managers report any issues to them. Final say when making store related purchases.
Frustrations
Surprise health-inspection deductions that threaten reputation, high staff turnover forcing constant training on store protocols.
When cold outreach failed, I went on-site.
Due to the constraints of this course final, I needed to get data fast, so I opted for 'guerilla' style interviews, where I would go around town and targeted appropriate restaurant demographics. This included some chains in my area and local mall. I ended up getting interviews with store managers at Chipotle, Subway, and a local bubble tea shop.
Research results
My Findings
In my notes, I found three main themes.
[Initial Research]
High turn over rates cripples training.
[Initial Research]
[Initial Research]
Managers can’t be everywhere.
[Initial Research]
[Initial Research]
Training is one and done.
[Initial Research]
My takeaways
The main themes allowed me map out core components of what a solution may look like, not only that, with current AI advancements, the ceiling for what is possible is higher than ever, which brought me to think about computer-vision technology and how it can assist business owners and employees handle food safety .
Relying on one role isn’t enough
Cameras monitor every prep station, backing up managers who simply can’t be everywhere at once.
Software should be familiar
With high turnover rates, software should be accessible and mirror other software already used in the workplace (i.e POS, Food Delivery Apps, Resy, etc).
Information should be clearly accessible
Procedures, solutions, and ingredients should be readily available so that action can be taken, without the assistance of a manager.
How might we help bar-style crew catch and correct contamination before the next order leaves the line?
Ideation & prototyping
My first challenge was deciding how I wanted the interface to be formatted. Due to the nature of the software, it would most likely be used on a tablet of some sort, with elements big enough to be legible from a medium distance, while also being glove friendly.
Looking at how other software optimizes for the same/similar user groups got me started fast, what I found was that a dual pane format was common and easy to learn.
Interestingly, looking at home security software was also helpful.
Lo-fi's
My first challenge was deciding how I wanted the interface to be formatted. Due to the nature of the software, it would most likely be used on a tablet of some sort, with elements big enough to be legible from a medium distance, while also being glove friendly.
Current version
My first challenge was deciding how I wanted the interface to be formatted. Due to the nature of the software, it would most likely be used on a tablet of some sort, with elements big enough to be legible from a medium distance, while also being glove friendly.
Reflection
What I would do differently
This project reminded me that the best concepts are born in the field, not the inbox: when 22 cold emails went unanswered, face-to-face interviews at the mall uncovered the real pain lurking behind cross-contamination. Pivoting from an “allergen-only guard” to a broader food-safety co-pilot taught me to design for layered accountability— an interface any new hire can grasp in minutes, yet robust enough to satisfy auditors and protect the bottom line,