All things food
Cross-functional partners
Potential investors
AI based restaurant search and recommendations engine.
All Things Food
I founded and designed All Things Food, an AI restaurant-matching app that helps friends decide where to eat, and took it from an idea to a tested prototype in two months.
Audience: Pitched to potential investors.
All Things Food: designing an AI that decides where friends should eat
ResultTook an AI restaurant-matching idea from zero to a tested prototype that matches friends by taste.
The problem
Choosing where to eat with friends is hard when everyone’s tastes differ, and no app solved it for groups.
What I did
- Researched the problem through competitor studies and user interviews
- Designed AI taste-compatibility scores that show why each restaurant is suggested
- Designed group recommendations, group chat and a snapshot view, then tested them with users
How might we find restaurant recommendations for multiple friends?
Every time my friends and I planned a meal out, we got stuck on the same questions. Would they like it? What are they in the mood for? Who should I even invite?
Everyone in a group has different tastes, and no app helped friends decide together. Eating out with people who share your taste could be a game changer.
An app that shows which friends share your taste, recommends restaurants and cuisines your friend or group will like, and connects you with locals for cooking experiences.
Nothing like it existed, so I designed it from zero
All Things Food had no direct competitors. I explored two directions: a standalone app, or a feature inside an existing app.
No ‘before’ here.
This didn’t exist until I designed it.
- No app compared tastes between friends
- No direct competitors to learn from
- Groups were left guessing where to eat
to a shared
taste match
- An AI restaurant search and recommendations engine
- Taste-match scores between friends
- Group chat, interactive maps and a snapshot view
I asked people before designing anything
With no competitors to copy, research carried the project. I studied food delivery and travel apps, ran user interviews with prepared questions, and polled people on Instagram.
73%find it hard to discover new places to eat when they visit somewhere new.
90%would use an app that shows which friends share their taste.
What I set out to learn
- Who it’s for. The target user group and how they plan meals out.
- What makes it different. The key differentiators for the app.
- What people need. Their pain points, needs and priorities.
Four ways I pressure-tested the concept
It was a self-sponsored project, so I ran every step myself, from research to critiques.
How the concept got sharper
- Competitor analysis. With no direct competitors, I studied food delivery and travel apps.
- User research. I planned and ran all of the research myself.
- Product reviews. I gathered lots of feedback from the target audience.
- Design critiques. I presented the concept to people across the industry for as much feedback as possible.
What usability testing changed
The biggest insights came after research and testing. These are the changes they led to.
Prototype walkthrough. Profile, taste matches and top cuisines.
Group chat moved to the nav bar
It first lived inside the profile. After usability testing, the inbox got its own spot in the bottom navigation bar, one tap away.
The map became interactive
It started as a static view, like top cuisines and ratings. Testing showed that making it interactive gave the app its “aha!” moment.
It can also be a feature
The concept works as its own app, and it could also plug into an app or platform that already exists.
Three flows that help friends decide together
Picked from many iterations across the whole platform: its screens, their variations and the new features it needed.
Find friends who eat like you
See which friends share your taste, and the top cuisines you both like.
Pick a place as a group
The same match for a whole group: the top cuisines everyone will enjoy.
Compare tastes side by side
Compare your taste with a friend’s and get cuisines and restaurants you’ll both like.
Still cooking, and growing
Implementation is underway, with growth experiments including A/B testing.
Pitching it as a feature to bigger brands.
The app is in development now.
The full case study deck is available on request ☺
Objective
How might we find restaurant recommendations for multiple friends?
Overview
Context
The thought for this project came up when my friends and I were deciding where to go out and eat. Usually, when we go out to eat with friends we think about where we should go. Would he/she like it? What kind of food would they prefer? Whom can I go with?
Different people have different preferences and dining together could be a game changer for many. This is where the idea of all things food came in.
Doesn’t exist
Outcome
This app shows you which friend has a similar taste as yours, restaurant recommendations, cuisines your friend or group would like, and cooking experiences you can have with locals.
Sneak peek
Before
User research
After
AI based restaurant search and recommendations engine
Before
At the time, there were no other features/apps with this technology in the consumer market. Hence, I relied heavily on user research:
1) Studying competitors in the market and how they approach the problem statement.
2) User interviews to have fluid conversations that include a set of prepared questions to understand the target user group and the difficulties they face.
My focus was on understanding the target user group for the application, Identifying the key differentiators in the app, & understanding the user's pain points, their needs, and priorities.
All Things Food offers a very unique solution and does not have direct competitors. This led us to explore two major directions: the app as a separate idea & it being incorporated as a feature in an existing app/company.
Design process
-
No direct competitors so studied food delivery and travel apps.
-
Since it was a self-sponsored project I was responsible for the user research as well.
-
Took a lot of user feedback from the target audience.
-
Presented the concept to a lot of people in the industry to get as much feedback as possible.
After
Final designs
Here are some of my iterations. This extensive project required me to analyze the entire platform experience, its post variations, and new features that needed to be incorporated.
User profile
Group chat
Similarly this shows the top cuisine for a group of friends.
Snapshot view
Ongoing.
A mobile application that is a restaurant search and recommendations engine that uses advanced machine learning and AI technology.
This allowed you to find friends who have a similar taste as you & what are the top cuisines you both might like.
This feature compares your taste with a friend & recommends cuisines/restaurants you both might like.
Results & learning
Implementation underway with growth experiments, including A/B testing.
Pitching it as a feature to bigger brands.
App development under progress.
The full case study deck is available on request. ☺
Challenges
Since it was a self-sponsored project there were a couple of challenges/insights that arrived after user research and user testing.
User testing
Since it was a self-sponsored project there were a couple of challenges/insights that arrived after user research and user testing.
-
At first, the group chat was designed to be accessed from the profile but after the usability testing, we decided to have direct access to the inbox from the bottom navigation bar.
-
At first, the map was designed in a similar way as the top cuisines and ratings but after the testing, it was concluded that adding interactivity to it would it the "aha!" effect.
-
This could also be an additional feature in an app/platform existing in the market.