Marrying AI and neuroscience to improve recommendation services

By January 20, 2015
Nara Logics

Nara Logics specialises in artificial intelligence and in analysing how the brain works as a basis for its personalised recommendation systems for companies.

The challenge that the Nara Logics founders set themselves when they launched the company back in 2010 was to understand how the brain works in order to be able to offer closely-tailored personalised recommendations. The Cambridge, Massachusetts-based startup, which specialises in artificial intelligence (AI), has neuroscientists from MIT and machine learning experts on the team. In 2012, the startup launched its recommendation and personalisation customer application platform, which goes somewhat further that the simple recommendation services provided by such firms as Yelp and Foursquare and is designed to help you find a restaurant or hotel that suits your recorded tastes.

The company’s new president Jana Eggers, whom L’Atelier met up with during the Lean Startup Conference in San Francisco in mid-December, told us about Nara Logics’ future plans. Following the initial platform targeted at end consumers, the firm is now pivoting more towards licensing its recommendation engine and underlying AI software to businesses. In fact, ever since was launched, corporates have been asking the company to design a platform specifically for their business sector. So Nara Logics is now shifting into B2B mode.

Targeted offering based on neuroscience

Today Nara Logics offers unique recommendation technology to a range of companies in the payments, e-commerce and tourism fields. The startup’s algorithm is able to analyse companies’ ‘big data’ and establish logical connections that create the link between what consumers want and what companies have to offer. However, it is vital that Nara Logic’s corporate clients maintain a constant flow of analysable data so that the workings of the brain and the various connections it makes can be transposed to best advantage, explained Jana Eggers.

In the end all the data is gathered together on a knowledge graphic for each company, highlighting the ‘digital DNA’ of its customers, i.e. their preferences in terms of search and personalisation. “If you have a preference for a particular restaurant but your friend has a different preference, s/he will receive different recommendations from your because your digital DNA just isn’t the same,” points out the Nara Logics president.  This type of analysis will give companies a better handle on their customers’ behaviour. The novel feature of the Nara algorithm is that it seeks to capture the interactions influencing customer behaviour from a neuroscience viewpoint.

Neuroscience requires more research

As technologies such as deep learning, image recognition and predictive APIs become more widespread, they are boosting the capacity of computers and IT systems to seriously disrupt all kinds of industries. AI is also a technology that is gradually revolutionising industry. “A lot of research has been done into the way machines work, but very few firms have looked into the way the brain works. That’s where neuroscience comes into what we’re doing,” explained Eggers, adding that the goal of Nara Logics is to bring real intelligence into artificial intelligence.

However, as yet there are very few neuroscientists working in business. Eggers points out that neuroscience is still quite a young field of enquiry and that there is not as yet enough documentation on the workings of the brain to be able to create a fully neuroscience-based approach to AI. The algorithm that Nara Logics uses is based on studies published just a few years ago. Going forward, the company intends to improve its technology by working with people from different fields who have already tried other AI systems unsuccessfully and are now ready to test the new personalised recommendation technology.

Legal mentions © L’Atelier BNP Paribas