Indrek Vainu

Using neural networks for enhancing video call quality

One of the things our tech team is always fascinated by is the way in which neural networks can be deployed to solve a variety of use cases. And even though we work with conversational text and chatbots to improve customer service for our customers, we are inherently curious and tackle various challenges through our experimental work. So it is no wonder that a few months ago, during one of our casual Friday afternoon brainstorming sessions, our talk drifted to the poor quality of video calls with an iPhone. It is well known that the popularity in various entertainment and

Read More »

What We Learned At #dmexco – the Marketing/Advertising Expo

Customers are at the center of every business. Keeping customers happy is what makes the business grow. While simple in its essence, customer service is often an overlooked aspect. For many it is a thing that would be best kept at a comfortable distance – to outsource it to call centers abroad and maintain a variety of email/web forms through which customers can interact. Quite often one might wonder whether the crafted email will reach a service representative or end up in a black hole. We just came back from Cologne where we attended dmexco, Europe largest marketing and advertising

Read More »

3 Elements That Make Bots Work In Facebook Messenger

The brave new world of bots is marching onward with each passing day to the steady drumbeat of growth in messaging apps worldwide. People love to text and messaging has emerged as the preferred communication mode for many. With the growth in Facebook Messenger and its platform, talking to companies and brands through messaging is slowly starting to gain ground. As with any breakout thing, there are opinions on both sides of the spectrum of whether it is hype or whether it is something tangible that will stick around. Whatever the case may be (I’ll leave my crystal ball post

Read More »

Fashion brand’s Messenger store launch covered in US media

CartSkill builds one of Europe’s first online stores in Facebook Messenger for fashion brand Monton. Now users can shop inside Messenger by chatting with the Monton brand. Instructions here. “We think of this as part of an omnichannel approach,” said Külli Koort, head of ecommerce at Monton, which operates an ecommerce site and just shy of 100 brick-and-mortar locations across eastern Europe. The idea is to create an online store within Facebook Messenger, piggybacking off Facebook’s existing app with mass adoption, while at the same time reducing the number of clicks it takes to get people ensconced in Monton’s ecommerce

Read More »

Deep learning networks benchmarked.

As we have made a good number of improvements in our visual recommendation technology we decided to benchmark it against other popular deep learning networks. The aim was to compare our recommendation against AlexNet (Caffe implementation) and ResNet (Torch implementation) and to outline errors in product similarity identification. The setup for the comparison was such that with each network we fed in 3 different product categories and identified visually similar products based on a single thumbnail image. The errors are outlined in the images below in red circles. For the product categories we used 3260 different bags, 7028 different shoes

Read More »

What’s in a message – recap from Latitude59 and customer reactions

Last week we were selected as one of the top 10 early stage start-ups in Latitude59, one of the largest start-up conferences in the Nordics. Aside from pitching in front of a packed demo room, we were fortunate to interact with dozens of people in our demo booth. This provided an especially cool insight into how people react to bots and what are the main questions they have. Below is a list of things that came up when discussing building ecommerce stores and customer service solutions into Facebook Messenger: “Wow, I did not know that was possible” – this was

Read More »

It’s A Bot World

A few weeks ago Facebook announced that developers can start building bots for their Messenger platform. Today we’re happy to announce that we have released our shopping bot for Messenger by harnessing our artificial intelligence engine in e-commerce. It is a nice preview of things to come. As we’ve been focused deeply on Messenger and the bot world, I thought I’d share a few thoughts around the topic. Why is this significant? Bots are a new type of interface for interacting with businesses, especially on messaging platforms. This is important as the top messaging apps have more monthly users than

Read More »

What Media Companies Are Looking For

We at CartSkill recently attended the media and start-up matching event MediaHoneypot in Helsinki put together by the good people at ArcticStartup. As we’ve been deeply engaged in the world of media, advertising, marketing and publishing over the past months, it was a good chance to get some fresh perspective from a number of large media companies both in Scandinavia and the rest of Europe. A couple of messages that we picked up: Ad blocking is (still) a problem. Ads pay for the content but blocking them means that producing that content will get more expensive. So media companies are developing different

Read More »

Hate Banner Ads? We’ve Got the Cure.

In our previous post we covered how our visual product discovery system solves the problem of product discovery. In this post we focus on how the same technology is used to solve the problem of “banner blindness” — a case where the visitor to a website chooses to ignore the banner ads. I guess it needs no further explanation, we’ve all ignored banner ads. They flash, play tedious videos, load slowly and show random products. They are mostly irrelevant. Yet, the content on the web would be different if it were not for banner ads and other type of digital advertising. Much

Read More »

Using the Power of Our Artificial Intelligence in Discovering New Products

Recently there has been much talk about Google’s AlphaGo that defeatedthe best human player in the game of Go. A remarkable step that shows the potential of artificial intelligence. Neil Lawrence from The Guardian had an especially interesting take on the subject as he drew a parallel with the development of the steam engine in 1712. The steam engine was initially built specifically for a single use case, pumping water out of coal mines. However, it was James Watt who made significant improvements in the efficiency of the steam engine’s condenser so that it could be used in many other fields

Read More »

Subscribe to Our Newsletter