TOKEN
2023
Leveraging AI in my design process to craft a home dashboard and customer threats screen for Token
At Chase, some of my projects included working on the rewards mobile gift card flow and refining the navigation of this space. These two shorter snapshots of my time at Chase are shared here.
Role
Product designer
Timeline
1 year
Team
Senior designers, PMs, engineers, content designers, marketers
CONTEXT
Dispensaries face many threats that they need to guard against
The customer "risk" flow would sit within the main dashboard, where order creation took place
The home dashboard featured analytics, inventory information, and transaction details
The transactions tab would have filters allowing merchants to identify flagged risks by threat level
Risk alerts would sit at different points in the process, adding friction based on severity
DESIGN PROCESS
I used a combination of conversational tools (Chat GPT, Gemini) and design tools (Gemini, v0, Bolt, Lovable)
FIG 2.1
USING DIFFERENT TOOLS FOR DIFFERENT PURPOSES
1
I leveraged LLMs to synthesize the initial research
I asked the LLM questions about research themes from surveys and interviews as well as how to logically structure flows
FIG 2.2
QUESTION'S I'VE ASKED CHATGPT
2
I used generative design tools (v0, Bolt, etc) to explore ideas for key screens
I then analyzed the output to find commonalities and drawbacks
FIG 2.3
ANALYZING INSIGHTS FROM BOLT/LOVABLE DESIGN IDEAS
3
I fed my designs into Gemini & v0 for a quick working prototype
This step is just to get a sense of how all the moving pieces fit together as a full flow, and fits in after I have made a few iterations of my own.
FIG 2.4
EXPLORING PROTOTYPED VERSIONS TO ITERATE UPON
4
I then evaluated the effectiveness of various versions with Chat GPT
By asking pointed questions like, “Does version 1 or version 2 more effectively reduce user friction?”, I was able to get answers that pointed me in the right direction.
FIG 2.5
ASKING CHATGPT FOR FEEDBACK DURING THE PROCESS
DESIGN PROCESS
Throughout this process, I was iterating on what I had learned
AI became a way to speed up my understanding and iterating—not take it over completely
This is what my workspace looked like during during steps 2-3, prior to step 4 (where I received feedback from ChatGPT).
FIG 3.1
MY OWN EXPLORATIONS IN FIGMA
SOLUTION
Token: Customer Risk Flow
FIG 4.1
home dashboard, transactions tab, and slide in panels
