<aside> ℹ️

As an engineer at Listen, we expect you to take a project from idea to completion. This exercise is similar to what you would do before implementing a new project. We will evaluate your product thinking and technical architecture and want to feel confident that you could take a project like this to success.

</aside>

Ask-me-anything Analysis

For each study we already generate a static analysis report. However, researchers often have specific questions they are interested in that might not be answered in the report.

So we want implement a new feature for our analysis, the ask-me-anything function: a textbox where users can be able to ask questions dynamically and get a mind-blowing answer.

Note that this can come in very different forms, from a quick short response to an elaborate report. You pick the direction for this but our customers are probably most interested in a sophisticated report rather than a vague answer.

Please Prepare

• For inspiration, it might help to look at what adjacent players are doing (ChatGPT, Perplexity, Hebbia) even though their use case will be different.
• Please present your thoughts in a Google doc or similar.
• If it helps, you can sketch out the interaction flow in Figma/Figjam
• If you have questions, please reach out via email or ask within the session. We have more context on the problem, please utilize that like you would if you actually worked here.

Depending on your familiarity, we recommend spending one evening (max 2-3 hours) on this.

Who’s the customer

• In this case it's a CMO at Five Below, a value retailer with over 1,300 stores. Remember to keep that in mind when you present your thoughts.
• Think about what problems they might have. They probably need to present insights and convince other stakeholders.
• They might need to figure our how to position themself against Temu / Shein.

You should start by answering these questions

• What are the pain points we solve with this feature for this specific customer? How do they generalize to other customers?
• How would our stakeholder measure success of this feature?