This is the search tool that you have been waiting for, the one that actually understands what you mean when looking for products on an e-commerce website
Israeli search revolutionaries Twiggle announced on Thursday that they have closed their Series A funding round, emerging with $12.5 million in new financing to change the way that people perform e-commerce searches online completely.
The round was led by internet and media investor giant Naspers, whose portfolio includes companies like Tencent, mail.ru, Flipkart, SimilarWeb, and Souq.com. Also participating in the round were Yahoo Japan, State of Mind Ventures, and Sir Ronald Cohen.
Twiggle was co-founded in December of 2013 by CEO Dr. Amir Konigsberg, previously one of the members of Google’s emerging markets operations and former managing director of MySupermarket.com, and Dr. Adi Avidor, a former engineering tech lead at Google.
The two have authored over 35 U.S. patents and bring a wealth of experience in digital innovation in the fields of Search, AI, and e-commerce.
Changing the way that humans interact with search engines
It should come as no surprise that search engines and humans speak different languages. Whereas people (in most cases) understand how to use context to draw out the relevant pieces of information from a string of words and put them in order to perform the most relevant response, search engines will simply look at the words and make its own inferences on what is important, often without any regard to the context.
One of the places where most people run into this human / digital divide is when it comes to shopping online. Let’s say that you are looking for a new smartphone and having a strong battery is very important.
A reasonable search query would be to type, “smartphone with the best battery”. Unfortunately that string of words using some of the most popular places to shop online is unlikely to lead you to anywhere close to what you were looking to find.
Here’s what you find on Amazon;
Amazon’s search was probably the best of the worst, since at least it showed external batteries for smartphones in its top options. Walmart and B&H were so far off that it was embarrassing.
Google does a little bit better by pointing users to websites talking about smartphone batteries.
In all these cases, the end result is unsatisfactory and leads the user on a long and convoluted path unlikely to lead to a quick sales.
“Search is about people,” Konigsberg tells Geektime. “It’s about giving people the information they want and need. And it’s about giving this information in a way that people are able to consume it, so that they can form beliefs and make decisions based on it.”
The Twiggle team has built a search technology that works to bridge the human / machine understanding gap for the e-commerce space. They have developed a deep and encompassing data infrastructure uniquely built for digital commerce.
Essentially, the wealth of information on the internet simply has not been well organized. It sits there on Amazon and others’ servers in categories but fails to deliver actionable information.
What Twiggle does is collects and analyzes data about products and turns that data into readily available information that can then be used by consumers whenever they need it. Twiggle’s systems have achieved in-depth knowledge of products by understanding the world of e-commerce, and use this comprehension to turn data into information.
“What is unique about this infrastructure is that it reflects the totality of what we call ‘commerce knowledge’ available on the internet,” explains Konigsberg. “What we mean by knowledge is a structured mapping of all the billions of products and all attributes, features, and specifications of all products, and the relations between them.”
They have combed through the internet, looking for products and understanding what are the attributes that make up the products out there to bring users search results that are head and shoulders above what is currently available.
“Twiggle is taking Search to the next level,” says Konigsberg. “Our engine is able to actually understand queries, understand what people mean and intend when then they type a query. And we are also able to provide results that respond directly to what people type.”
Twiggle goes a step further from just bringing the user more accurate search results, to incorporating Natural Language Processing (NLP) systems that can understand terms like thin, quiet, light, and other descriptors that a person would comprehend.
If for example you were looking for a “light laptop”, their search would know to look through different kinds of laptops, compare their weights, and finally bring you a selection of options that fall on the lower end of that spectrum.
Twiggle has built AI capabilities that can be used for interacting with users to help them narrow down their search. So after being presented with a group of lightweight laptops, the system can ask the users questions like the color, amount of memory, and other features that will bring them the most appropriate set of options to choose from.
They are currently working with a number of clients to integrate their solution into e-commerce platforms.
Funding will go to building out the engineering team, in deep learning, machine learning, and natural language processing, continue growing operations and the product. They also plan on branching out into new speech-to-text systems and conversational commerce systems.
Regarding competition, Konigsberg cites IBM, Oracle, and Adobe as other actors working in the field of search.
Moving into the future of search
The internet has grown to be the most extensive repository of information in human history. While the power to reach out to this immense resource for answers seems to be at the tip of our fingers, finding the information that we really want and what we’re looking for, technology has so far failed to adapt itself to interact properly with humans and how we behave. Basically the access to this information is being blocked by a jumbled search system.
At its core, technology should be shaping itself to meet human needs, and not people needing to learn how to perform a search engine-friendly query.
Twiggle appears to be at the forefront of reforming the internet to be that source of ready information, acting as a natural extension of ourselves.