Before a business plunges headfirst into AI, it is important to identify the use case – the exact problem you are trying to solve. In my experience it is vital to be as specific as you can. At mallowstreet, our use case was the meeting. 

We all have meetings – lots of them – probably too many. So, we started by asking ourselves the simple question: how can we help make meetings, and the follow-up, more efficient and effective? As a result, we built our own AI tool SOFI to analyse a meeting. 

We used our industry-specific knowledge to train SOFI to fluently summarise financial meetings and discussions, as well as surface key actions and takeaways. We also built the capability to analyse a meeting through SOFI’s lenses – covering risk, return, liquidity profile, investment time horizon, growth versus matching investment characteristics, and ESG. 

The result was amazing. As a meeting attendee, you could now actively listen and truly pay attention – feeling 100% present, knowing that SOFI is in the background ensuring no points from the conversation are missed.  

Then came our second round of development – again leaning on the use case approach. 

This time, we asked ourselves: if SOFI can do all of this for one meeting, can we apply SOFI analysis to a series of related meetings, to highlight key themes, identify discrepancies and surface common questions? This was the genesis of our Multi-Vertical Analysis (MVA). Multi-Vertical Analysis unlocked our trend analysis tools and paved the way for the development of ‘SOFI scores’, which is how we quantify key information. For example, how much time does each participant speak (soapbox score) or how many ‘ums’ and ‘ahs’ do you say (disfluency score)? With this layer, SOFI has also become a pitch/ presentation practice tool – providing feedback that is objective, consistent, and transparent. 

Our clients tell us SOFI is having a really positive impact on the way they work, and are applying their own individual user cases to SOFI. What’s your user case?