Video: What is the ‘Human API’ in capital markets?

Video transcript

The system and process and data complexity that typically persists within capital markets, a lot of the compensation for that manifests itself in people.

And when we look even at kind of small to medium-sized banks, it’s not uncommon for them to have 3, 4, 5,000 people sitting within their operations functions.

As you start to drill into what a lot of those people are doing, they are really plugging data and automation gaps in that infrastructure.

And a lot of what they’re actually doing is really compensating for poor data and a lack of automation.

We come to refer to a lot of those roles as effectively being the ‘Human API’, which really talks around the fact that the human needs to be sitting in between systems or processes: cleaning data, intervening, making decisions, perhaps on trades or data, or helping exceptions kind of through the process.

And we see these people and they’re easy to recognise when we start looking at shared inboxes being kind of really the workflow tool of choice for much of the capital markets industry.

You know, large teams of people whose role it is to sift through these inboxes, extract data that’s needed to drive some of the downstream post-trade processes: that could be trade matching; that could be kind of trade confirmation; that could be exchanging data to support settlement.

So, there’s a significant kind of volume of people who are associated with those relatively mundane or relatively manual tasks, but they are there purely to compensate for gaps in kind of data and process.

They persist further on down through the infrastructure, so they are quite often the people who may well be dual-keying between system A and system B. System A and B have perhaps different data models, and you can’t automate that interaction between them.

So the human API, sits between two of those systems, extracting data from one and uploading or manually inputting it into the other one.

They quite often also manifest themselves in the exchange between perhaps counterparties or market participants, where the language and the explanation of even common processes, such as trade confirmation or trade settlement, actually differ wildly from firm A to firm B.

And I think for us we see that everywhere. We see that in large organisations, quite sophisticated wholesale, universal banks right the way down to small asset managers and hedge funds.

And fundamentally it comes back to that point that actually, by being data centric and being able to start to explore causes of poor data, you will be able to unlock a lot of the activities that your Human APIs are doing. That enables you to automate them if you can, or perhaps unlocks additional value with them, where you’re able to deploy those people into higher value-added activities.