Conversational AI is a huge advancement. Massive. However, instead of blending it with existing workflows in creative ways, I have seen many companies just bolt on a chatbot or add an AI tab... In this case study I wanted to share a couple ideas I’ve had over the past month that I believe will resonate with the users’ Jobs To Be Done instead of distracting from the core experience.
A hands off CRM
Scout is a CRM tool for Venture Capital Investors (VCs)
In user interviews, four VCs expressed frustration that their CRMs were out of date. The common reason? A lack of time to manually input information between back-to-back calls. This highlighted a clear gap in the market: none of the existing CRM tools effectively solve this pain point. If Scout can address it, we have a strong opportunity to differentiate ourselves from established competitors.
I designed a concept to automate CRM entry using AI. When a VC joins a call with a transcript tool like Granola, the tool would sync with Scout to parse the transcript. Scout AI would then extract key details tied to specific entry fields and auto-populate them—creating a centralised, accurate source of truth across the organisation.
Video editor, at your command
Taka helps students in the US earn athletic scholarships by analysing their soccer games, breaking them into key moments, and turning those clips into compelling showreels.
Competitors such as Hudl use outdated manual video editors, which require a level of digital literacy and eat up a lot of time for kids who already have busy schedules. What if parents with minimal digital skills could effortlessly turn their kids’ soccer games into highlight reels to share with relatives or preserve as keepsakes? Imagine if young athletes could instantly generate reels to review their own performance seconds before practice, helping them improve game by game.
I designed an AI-powered tool that scans our video database, retrieves clips based on a user’s query, and automatically compiles them into a polished showreel—ready to share, review, or submit for scholarships.
A front page, just for you
The Hong Kong Jockey Club is a trusted provider of sports statistics for millions of patrons.
The stats page currently offers the same content to every user—whether they're tracking the Premier League or following the Hong Kong International Races. This one-size-fits-all approach misses the opportunity to engage users with personalised content. Could we drive more repeat visits by serving tailored information in a more interesting format?
I designed a dynamic dashboard powered by AI, featuring a library of bright and exciting data visualisation templates. These templates can be randomly chosen and populated with content personalised to each user's interests—turning static information into an engaging roulette designed to surprise and delight every time.
Find what you need instantly
Taka helps teams review past games and analyse their on-field performance, turning raw footage into actionable insights.
Opportunity
Originally, users had to rely on a set of manual filters to search through footage and athletes—The process felt more like using a spreadsheet than watching a game. Could we use AI to make this experience feel more intuitive and human?
Action
I proposed a shift to natural language search. Instead of clicking through dropdowns, users could simply type phrases like “show me my team” or “find the goal we conceded in our last game” The AI would then surface the relevant clips in our usual UI as opposed to a separate chat or canvas.
While most of my client reviews are NDA-protected (because, you know, top-secret agency white label stuff), I managed to sneak in a few favorites from my previous partners.
