Project Overview
01
Features
02
Objective
03
Insights & Testing
04
Design & Development
05
Download Resume
Contact
3.1
How might we
make small talk effortless/eliminate it?
let AI clones/personas do the small talk for users.
make connecting with someone not based on superficial reasons?
have a randomize button, don’t reveal full name.
make it a personalized and safe experience?
only allow verified college students be a part, allow voice cloning for AI.
4.1
Research & Testing Insights
My Role and development
We conducted multiple interviews, tree testing, open and closed card sorting, and cultural probes
to inform our design process and decide the features for the MVP.
Visually, we aimed to reflect the company's values by creating a modern yet vibrant aesthetic. The goal was to blend a sleek AI-inspired look with retro elements, ensuring it felt like a social media platform rather than a game
I was the sole/Founding Product Designer, leading the end-to-end design and development of ICE, an AI-driven app using voice cloning to help users meet new people.I led the product strategy, roadmap, planning, testing, and research. Though we didn’t secure funding, the app was featured in a New York Times article. https://www.nytimes.com/2024/11/14/technology/ai-dating-apps.html
05
Design & Development
© Sejal’s Portfolio
All Rights Reserved
Project Overview
01
Features
02
Objective
03
Insights
04
Design & Development
05
02
Features
Ice lets users create a unique AI Persona that reflects their story, interests, and personality through curated icebreakers. Instead of jumping straight into conversations, matches engage with your Persona first in a 20-message limit chat, helping both sides get a vibe check without pressure. You can review these chats, decide who you connect with, and “break the ice” to reveal your identity. From personality-first matchmaking to easing the anxiety of small talk, every feature is designed to make meeting people feel natural, fun, and low-pressure.
Onboarding
Explore Page
Chats
Profile
03
Objective
The idea originated as a social experiment after hearing countless complaints from people in New York about how hard it was to make friends or date in the city, with many frustrated by the challenge of small talk on dating apps.
Along with three other Columbia University grads, we built the MVP within 5 months from the idea to a fully developed app, testing with Columbia students and launching on the App Store.
3.1
How might we
make small talk effortless/eliminate it?
let AI clones/personas do the small talk for users.
make connecting with someone not based on superficial reasons?
have a randomize button, don’t reveal full name.
make it a personalized and safe experience?
only allow verified college students be a part, allow voice cloning for AI.
04
Insights
4.1
Research & Testing Insights
Our testing audience
With each of our testing phase, we released test flight on app store, testing with our friends for bugs, as well as went around Columbia recruiting users for the app and asking them to test for us. We tested with over 150 users.
Lo-fidelity testing
In the low-fi phase, I experimented with different concepts for the explore page, focusing on task flows. Initially, I considered a map-based UI, but given the need to launch the app quickly, we opted for a simpler swiping interface, which was faster and more intuitive. Made other interaction changes.
Mid-fidelity testing
I created a fun mid-fi design and tested it with Columbia grads, who responded positively to the visual aesthetics. However, we noticed that not displaying a picture of the person interacting with the bot made users uncomfortable.
Final MVP Adjustments
To address the discomfort, we decided to show a blurred image of the person for the final MVP. We adjusted some other interaction experiences. Feedback suggested that the design resembled a retro game, so I refined the visual elements to better align with the app’s intended vibe.
My Role and development
We conducted multiple interviews, tree testing, open and closed card sorting, and cultural probes
to inform our design process and decide the features for the MVP.
Visually, we aimed to reflect the company's values by creating a modern yet vibrant aesthetic. The goal was to blend a sleek AI-inspired look with retro elements, ensuring it felt like a social media platform rather than a game
I was the sole/Founding Product Designer, leading the end-to-end design and development of ICE, an AI-driven app using voice cloning to help users meet new people.I led the product strategy, roadmap, planning, testing, and research. Though we didn’t secure funding, the app was featured in a New York Times article. https://www.nytimes.com/2024/11/14/technology/ai-dating-apps.html
05
Design & Development
Reach out to learn more!
© Sejal’s Portfolio
All Rights Reserved
Project Overview
01
Features
02
Objective
03
Insights & Testing
04
Design & Development
05
Download Resume
Contact
02
Features
Ice lets users create a unique AI Persona that reflects their story, interests, and personality through curated icebreakers. Instead of jumping straight into conversations, matches engage with your Persona first in a 20-message limit chat, helping both sides get a vibe check without pressure. You can review these chats, decide who you connect with, and “break the ice” to reveal your identity. From personality-first matchmaking to easing the anxiety of small talk, every feature is designed to make meeting people feel natural, fun, and low-pressure.
Onboarding
Explore Page
Chats
Profile
03
Objective
The idea originated as a social experiment after hearing countless complaints from people in New York about how hard it was to make friends or date in the city, with many frustrated by the challenge of small talk on dating apps.
Along with three other Columbia University grads, we built the MVP within 5 months from the idea to a fully developed app, testing with Columbia students and launching on the App Store.
3.1
How might we
make small talk effortless/eliminate it?
let AI clones/personas do the small talk for users.
make connecting with someone not based on superficial reasons?
have a randomize button, don’t reveal full name.
make it a personalized and safe experience?
only allow verified college students be a part, allow voice cloning for AI.
04
Insights
4.1
Research & Testing Insights
Our testing audience
With each of our testing phase, we released test flight on app store, testing with our friends for bugs, as well as went around Columbia recruiting users for the app and asking them to test for us. We tested with over 150 users.
Lo-fidelity testing
In the low-fi phase, I experimented with different concepts for the explore page, focusing on task flows. Initially, I considered a map-based UI, but given the need to launch the app quickly, we opted for a simpler swiping interface, which was faster and more intuitive. Made other interaction changes.
Mid-fidelity testing
I created a fun mid-fi design and tested it with Columbia grads, who responded positively to the visual aesthetics. However, we noticed that not displaying a picture of the person interacting with the bot made users uncomfortable.
Final MVP Adjustments
To address the discomfort, we decided to show a blurred image of the person for the final MVP. We adjusted some other interaction experiences. Feedback suggested that the design resembled a retro game, so I refined the visual elements to better align with the app’s intended vibe.
My Role and development
We conducted multiple interviews, tree testing, open and closed card sorting, and cultural probes
to inform our design process and decide the features for the MVP.
Visually, we aimed to reflect the company's values by creating a modern yet vibrant aesthetic. The goal was to blend a sleek AI-inspired look with retro elements, ensuring it felt like a social media platform rather than a game
I was the sole/Founding Product Designer, leading the end-to-end design and development of ICE, an AI-driven app using voice cloning to help users meet new people.I led the product strategy, roadmap, planning, testing, and research. Though we didn’t secure funding, the app was featured in a New York Times article. https://www.nytimes.com/2024/11/14/technology/ai-dating-apps.html
05
Design & Development
Reach out to learn more!
© Sejal’s Portfolio
All Rights Reserved