Digital Reasoning – AI That Understands You
Table of contents
Compliance is another one of those bureaucratic departments like Human Resources or Legal that largely serve as a bit of CYA so the company directors can sleep well at night. In departments like compliance, any notion of efficiency or effectiveness is thrown out the window in favor of rules-based procedures that make absolutely no sense when applied to the real world. If an employee takes $1,000 and buys some shares that are on a “restricted list”, then the end result is several days’ worth of back-and-forth with HR, involvement of a few highly-paid MDs, and a “written warning”. Never mind it’s a $1,000 stock investment and completely irrelevant to the firm’s fiduciary obligations to their shareholders.
This bureaucratic mire of time-wasting that was once called compliance will soon be relegated to an entity with the half-a-brain it takes to recognize a real compliance problem – as opposed to some poorly paid company grunt trying desperately to pick the next Microsoft so he can escape hell. That’s where companies like Digital Reasoning come into play.
About Digital Reasoning
Founded in 2000, Tennessee startup Digital Reasoning has now taken in $104 million in funding from some big-name financial firms like Goldman Sachs, Barclays, BNP Paribas, NASDAQ, and Credit Suisse. We first came across Digital Reasoning in our article on Artificial Intelligence (AI) as a Service and “Core AI” in which we noted that their technology, Synthesys, is exceptionally good at understanding what people are actually saying when they use analogies. It’s like when you come across some really good coke and email your boys on the sales team to let them know there is “snow on the ground”. If you say that and it happens to be the middle of July, Synthesys would easily be able to understand that what you are saying makes no sense, and would flag this for follow up by some tool in compliance. Here’s an actual example from Digital Reasoning:
Looks like someone is doing some colluding in the foreign exchange market which is a big no-no. Let’s talk a bit more about using Synthesys for compliance.
Digital Reasoning for Compliance
The next time we came across Digital Reasoning was in our article on 9 Regulatory and Compliance Regtech Startups in which we noted that adopting their technology is like adding “the brainpower of thousands of people to your team”, and provided an example from the company in hard numbers:
Synthesys had delivered impressive results to institutions including UBS, Goldman Sachs, Point 72 and many others with documented outcomes such as a 250% increase in detection acuity, 50% reduction in false positives, and a quadrupling in efficiency making it feasible to extend electronic communications surveillance to 100% of employees.
It shouldn’t be news to anyone who works in a publicly traded corporation that every single piece of communication you send on firm equipment is being monitored. What should be news is that now we’re beyond keyword matching, so you can assume that the monitoring system is now able to understand when you tell your cow-orker that the regional MD is being a “fcuk1ng pr1ck”.
Let’s be clear that every single publicly-traded firm out there worries about regulatory bodies like the SEC, and consequently employs some form of protection now. That protection is just so inefficient that it can’t be used to cover all employees, just some. The below table compares the artificial intelligence algorithms from Digital Reasoning to some older forms of monitoring like pattern-matching or keyword-based solutions:
In the olden days, being increasingly watchful of employee communications would often lead to a bunch of false positives that would result in pitting “rules-based John-in-Mumbai” types against “increasingly frustrated front-office” types, leading to a complete waste of resources all around. With AI algorithms, false positives are reduced which means less manpower is needed in your compliance department. Just how much money can be saved here? According to Salary.com, your average compliance manager makes between $86,658-$121,244 with “top compliance executives” pulling down between $182,650–$269,622. Imagine how much budget a 50-person compliance department consumes.
One of the reasons Digital Reasoning excels at finding crooks is because it has roots in this type of work. Their Synthesys platform was first presented to the U.S. Army back in 2002, and used for whatever war in the Middle East the Americans were fighting at the time. It worked so well that in 2010, In-Q-Tel (the branch of the CIA dedicated to new technologies to protect ‘Murica), made an investment in Digital Reasoning. But this doesn’t mean that the software can’t be put to use in other verticals – like customer service.
Digital Reasoning for Customer Service
As you can imagine, the ability to understand what someone is saying is particularly useful when it comes to understanding what your customers are saying – across all channels:
With the advanced natural language processing capabilities of AI that are all but commonplace these days, it’s very easy to collect all the unstructured data exhaust from your customers and transform it into a real-time temperature reading of what your customers think about your products and brand. What’s even more interesting is this notion of “connecting the dots” when comparing internal data sources to external data sources. For example, you may be able to connect complaints on a social media account to an actual customer record in your database and suddenly know the entire background of the transaction that caused the complaint. It’s all about understanding what your customers are saying about you.
Want to hear the best part? Unlike most of your tenured executives, the Synthesys algorithms will only get better with time. They will also never retire or hit on the interns at the company Christmas party. Not to mention that the Synthesys algorithms can handle an incredible amount of information when compared to humans. Digital Reasoning put together a white paper with Oracle that showed their system is capable of reading and analyzing in a few hours, the same number of documents that a team of 300 people could analyze in a year. Let that sink in for a moment and then think about all the other possible applications this technology could have.
Digital Reasoning for Healthcare
Digital Reasoning is just barely scratching the surface when it comes to all the potential use cases for their technology, though they also appear to be looking towards the healthcare industry. Remember how we talked about electronic medical records (EMRs) before? Here’s what Digital Reasoning has to say about EMRs:
Electronic medical records (EMRs) were intended to improve documentation of patient journeys, but in practice they have often disrupted workflows, been a source of inaccurate data, and added a barrier to the doctor-patient relationship
You can probably guess what comes next. The solution proposed by Digital Reasoning transforms the paper-based processes that are already in place “using proprietary Mimeo technology”, into a digital format so that they can be fed to Synthesys which will then place all that data into the appropriate EMRs. You can imagine how important it is to understand what the doctor is saying in freeform text fields, so that’s where Synthesys can excel. At the end of the day, the doctor can continue to use the old forms while also using EMRs that will soon be required for all patients. Everyone wins. Well, maybe not everyone. The Synthesys platform seems to offer such an advantage that inevitably there will be some losers out there – like CA Inc.
A Word of Warning to CA Inc.
Investors in the $14 billion computer software firm CA Inc. (CA) should take a look at the Digital Reasoning homepage right now. The very first thing they will see is some marketing collateral around migrating existing clients of CA onto the Synthesis platform. That’s a pretty strong message being sent to CA about how their CA Data Protection tool is an easy target for competitors like Digital Reasoning. More importantly, it shows how AI startups with powerful cognitive computing capabilities will cannibalize all sorts of existing software products by showing exemplary performance beyond the bounds of traditional software capabilities. Hard numbers about costs savings make for a very easy sell to the C-Suite types these days.
The power of technologies like Synthesys is that soon we will be able to create a clever AI chatbot that will be indistinguishable from humans. This is a tipping point because in today’s large corporations, most people you interact with on the job you will never meet face-to-face. It’s only a matter of time before many jobs are regulated to chatbot entities that live in the servers and interact with humans in a manner that is indistinguishable from an actual human. It’s only a matter of time before all these rules-based bureaucratic departments get replaced by chatbots – and we owe that pleasure to companies like Digital Reasoning.
Share
