Charles Post, Executive Vice President at Cimplifi, sees legal technology as bringing an end to ‘human APIs’ and ‘protein processing’, i.e. to take the person out of the data transfer process, especially in relation to contracts.
Taking humans out of mundane work and where doing so improves results is of course not just the dream of legal technologists, but arguably was one of the key aspects of the industrial revolution and then the digital revolution that followed it. The question now is: how far can we take it?
First, some context, Artificial Lawyer was speaking to Post (pictured above) primarily about Cimplifi’s work with derivatives: complex financial instruments based on future outcomes that are accompanied with even more complex documentation detailing dozens of ‘what ifs’ related to the deal. Such agreements cover, to put it simply, how various parties will pay each other in future scenarios, how risky different scenarios are, and if all the parties involved will be good for the money they might owe depending on those outcomes.
[And if you’d like to see a beginner’s guide to derivatives, check out this page from Investopedia.]
As Post said, some derivatives agreements can be negotiated over many months and may total hundreds of pages at the most complex end, for example when a giant corporation wants to hedge the future costs of its fuel needs and so in effect makes a bet, with perhaps multiple conditions on what various outcomes will mean for them and the counter-parties that accept the other end of ‘the bet’.
These contracts can be worth billions of dollars in some cases and as Post noted, they usually involve a lot of manual data entry as different players in the contract’s creation, dissemination and storage, move things along from stage to stage of the contract’s negotiation, final agreement, and then the recording of the key financial data based on the terms that were agreed. Now consider how these documents – which are in effect a massive, very complex ‘betting slip’ worth many millions – pass between multiple parties with people engaged at each stage to extract that information and input it into other systems, at a bank for example.
‘If someone gets a data entry wrong, such as someone inputs a currency exchange the wrong way, then this could cause tens of millions of dollars of damage. There are a lot of risks involved [with completing these contracts],’ Post explained.
Understandably, banks and corporates using them want to move more efficiently. Now here is the challenge: do everything manually and rely wholly on ‘human APIs’ to connect the various stages and use people’s brains as ‘protein processors’ to manage the co-ordination of masses of very complex data, or take a more tech-based approach and let the software – mostly – handle things. The human way inevitably is slow and also, despite people’s best efforts, can lead to data entry faults because this is the kind of work where mistakes happen.
‘This is a huge opportunity,’ said Post, ‘we can help with the automation of the contracts, we can provide checklists, we can get the data out that is embedded in the contracts, and track it to where it is consumed downstream.’
‘We can solve the problem [of manual data transfer],’ he added.
That said, there still needs to be some human input.
When it comes to building the complex automated documents for the clients that are running these derivative agreements there needs to be some very high-level human input to design them and make sure they are all set before being put into use. Cimplifi also offers human-assisted NLP analysis of existing contracts to help extract key data in previous agreements.
However, the key point remains: if Cimplifi and others, such as CreateiQ at Linklaters, can help not just with things such as the creation of what are very complex templates for contracts, but remove the need for ‘human APIs’ to capture that data and then pass it onto other parties for ‘downstream’ use without the need for ‘protein processing’ by specialised contract managers, then this could 1) speed up the entire process, 2) reduce risk as there will be a ‘hard-wired’ system in place, and 3) provide greater data transparency because with a more digital approach then access to a dashboard of the terms of a contract will be more easily visible and also easier to share with all the parties involved.
Removing humans from the process entirely seems unlikely, as someone still has to do the doc creation and negotiation, as noted. But, the connectivity between the contract, its data, and the parties involved…? Yes, that does indeed appear to very open to full automation and the removal of human APIs.
One last question: why is Cimplifi, which started off focused on eDiscovery, doing this? Although the automation approach noted above is not primarily about deploying NLP / text analytics software, the fundamental goal is the same: applying technology to text-based challenges in order to reduce the quantity of human labour in that process, and also gain the above noted benefits arriving as well.
To Post and colleagues, the movement from eDiscovery to helping with contracts (whether via automation or post-signature NLP analysis) is all part of the same logical continuum, and this site has to agree, this is the direction of travel, with human effort being pushed upwards rather than trapped inside monotonous processes.