Quantox Technology se priključuje misiji prikupljanja novčanih sredstava u okviru donatorske večeri koju organizuje humanitana organizacija “Srbi za Srbe”.

Donatorsko veče biće održano 20. juna u hotelu Royal Residence u Čačku sa ciljem pomoći porodici Bešanski, kao i Narodnoj kuhinji na Kosovu i Metohiji.

Domaćin dogadjaja biće proslavljeni košarkaš Srbije Aleksa Avramović, a sve prisutne očekuje bogat program! Cena ulaznice je ujedno i donacija - minimalne vrednosti od  100 eura po osobi. Pozivamo sve zainteresovane da učestvuju u sakupljanju 50.000 eura i doprinesu ovim humanim ciljevima.

Humanitarna organizacija “Srbi za Srbe” u proteklih 18 godina prikupila je više od 11 miliona evra i pomogla više od 3 hiljade porodica. 

‍

Pridružite se akciji i podržite one kojima je pomoć najpotrebnija!

Prijave za donorsko veče poslati na e-mail adresu: donatorskovece@srbizasrbe.org 

‍

‍

Više informacija je dostupno putem linka: https://drive.google.com/drive/folders/1dCaiaLw3YY4K7l81y4nIeEzoDfIsp2Or?usp=sharing

WHAT A GAME!

After 4 exciting days and a fair play fight - we have the champion of HIT CS:GO Vol.4!

Traditionally held in April, the HIT CS:GO tournament gathered the IT community again for the 4th time in a row.

The number of registered teams, companies and fund donations exceeded the previous years - over 40 teams from more than 35 companies competed in the famous Counter Strike: Global Offensive game!

In 2023, the tournament was dedicated to the youngest, with the goal to improve conditions in maternity hospitals and together we collected 2,291,900 RSD!

HIT 2023 WINNERS:

🥇 Quantox Technology

🥈 Vega IT

🥉 Levi9

We spoke to the winning Quantox team - Miloš Aksentijević, Marko Mitić, Luka Petrović, David Simonović, and Dušan Savanović - for the BTS of becoming a HIT CS:GO champion.

What are your impressions of the HIT Vol.4 tournament? How would you describe the atmosphere this year compared to previous years?

The whole team is united as much as coordinated!

Miloš Aksentijević / Ax.

The impressions are very positive, and the first place contributes a lot to that feeling. The idea is great and I would like it to expand so more teams and donations are coming from year to year.

Luka Petrović / hJk-.^

As in previous years - very positive impressions, a noble goal, and good banter.

Dušan Savanović / dukizen

The impressions are phenomenal! First of all, we love CS:GO, and especially the fact that this is the same team that has been playing and hanging out together for years. Also, the support from colleagues and the entire company truly surprised us and it was nice to see cheering in such a large number!

Who was the most challenging opponent during the competition?

The score shows!

Milos Aksentijević / Ax.

I couldn't decide on one team, so I'll say Levi9 and Vega IT. 

Levi9 seemed to have people who are very good at shooting, maybe even better than Vega. However, Vega was a tactically more prepared team.

Marko Mitić / MiTke-_o

Levi9 and VegaIT definitely. We won Levi9 with a score 2:0, but both maps were very close, and VegaIT took one map from us. If I have to choose one opponent - it would be Levi9.

Dušan Savanović / dukizen

From my perspective, definitely Levi9. Although we met them in the semi-finals, it was clear that the game was played at a much higher level, and this was shown on the first map, Vertigo, where we played as many as three overtimes. The boys from Levi showed a really competitive spirit and we had a really good match.‍

‍

We witnessed a very tense final of the tournament. Were there any decisive moments in the match and which tactics made the difference?

Maps & tactics - a winning combination!

Miloš Aksentijević / Ax.

Maybe there weren't so many decisive moments, but there were a lot of different tactics from our side. The idea was to occupy as much space as possible on the map, play more aggressively as CT, and make faster rotations. 

Luka Petrović / hJk-.^

I wouldn't say that there were any decisive moments. On our map peak and on the third map we played superbly, and the tactic was not to give them space to play their game - to attack and occupy space.

David Simonović / Unik ACM

There was tension in the final - as in every game. The opposing team led in some moments, where various factors had an effect. Primarily, it was a drop in concentration, which had the greatest impact during the game, and further, there have been short breaks between the semi-final and final matches.

Did you expect to win?

Expectations are expectations - and the result is in the end!

Marko Mitić / MiTke-_o

Honestly, yes... we knew it wouldn't be easy, but we were sure that we could win the tournament.

Luka Petrović / hJk-.^

Yes, because when it was announced that HIT will be organized again this year, we started to prepare and gather 2-3 times a week and play.

Dušan Savanović  / dukizen

I definitely didn't expect us to win and I can't say that my teammates would agree with me here. Not because of a lack of self-confidence, but because of the very fact that every year the tournament has attracted very good teams and players.

We believe that in addition to the tense fight, there were also plenty of laughs. Share some interesting anecdotes from the tournament :)

Enjoy the game!

Miloš Aksentijević  / Ax.

There was a lot of laughter and cheering, which is probably the key to our success. At no point did I feel like we were losing and that we didn't have a solution, so this made us play a very calm game at the top level.

David Simonović / Unik ACM

The whole tournament passed in a relaxed rhythm without tension, and that is the purpose - to enjoy the game, and victory comes just as a bonus!

Dušan Savanović  / dukizen

There was a lot of laughter, but I will point out one anecdote. Our teammate and colleague Mitke (Marko Mitić) is often overrun by planes when we play. It's not because he lives next to the airport - but because he plays the game on his laptop, which gets quite loud after a while.

So we persuaded him to put a “push to talk“ button to avoid those loud noises. During the finale, at one point, Mitke forgot to release that button, so we just heard Luka shout: MITKE AIRPLANE! 

And we really had a good laugh!

Laughter is a cure. Good competition is healthy. Enjoy the game. Help the community.

We are proud that HIT lasts and unites.

See you next year!

‍

AW Summit at Bucharest is a merge of learning, innovation and networking!

‍

More than 300 companies from different industries and from 80 countries will meet in Bucharest from 7th to 9th of May with the common goal to network, learn and share their practice.

As the largest European hub for affiliate marketers, digital companies, monetization platforms, entrepreneurs, content creators and social media specialists - the AW Summit conference gathers all those who are experienced in their business fields, willing to exchange ideas and build new business relations. 

With excellent keynote speeches, workshops, expo, and networking events, we are proud to be part of such a event with Quantox C level crew - Vuk Popović/ Founder, Filip Karaičić/ CEO, Danijel Petrović/ COO and Marko Ristanović/ Resource Manager.

As a leading software development company with a diverse team of tech professionals, we are creating top-notch software solutions and offer all-around IT support.

‍

Location: Piața Presei Libere 3-5, Bucharest

Booth#: 2

‍

We are waiting to meet you in person in Bucharest and discuss the latest IT topics and possible ways of successful collaboration!

‍

‍

‍

Quantox supports one of the oldest sports events in Serbia!

Just like in 2022, Quantox continues to support one of the oldest sports and recreational events that promotes healthy life and good habits for 46 years - the Fruška Gora Trail!

The oldest mountain and ultra-running marathon in Europe this year will be held on 22nd and 23rd of April. 

What does the Fruška Gora Trail stand for?

Quality time in nature and nurturing health habits!

Through 14 tracks of different lengths and height differences, the Fruška Gora Trail offers the chance to to experience all the beauty of the Fruška Gora National Park, as a recreationalist or as a experienced professional trail runner. 

https://www.fruskogorski-maraton.com/en/marathon-trails/

‍

When does Fruška Gora Trail start?

The official opening of the event is scheduled from 08:45 to 08:55 on the 22nd of April, and at 9:00 - the 46th Fruška Gora Trail will officially begin!

Sunday, April 23 from 11 am, is reserved for the youngest! Only children of preschool age accompanied by parents or other authorized adults can participate in the Path of Joy and Pleasure.

During a hiking tour or competitive race, all participants can feel the natural beauty of Fruška Gora - forests, streams, waterfalls, fresh air - but also, this is an opportunity to get to know the cultural and historical heritage. Along the marathon paths, you can see numerous Serbian medieval monasteries - Grgeteg (XV c.), Staro Hopovo (XVI c.), Novo Hopovo (XIV c.), Jazak (XVI c.), Bešenovo (VIII c.) Fruška Gora hamlets and tourist facilities.

Do not miss the unique opportunity to spend this weekend actively in nature!

After running kilometers and successfully completing the race - we are waiting for you to take a break in the Quantox Chill Out zone!

See you on the Fruška Gora!

‍

QUANTOX IT League is the first sports competition of this type in Serbia!

A motivation to enable people in this industry to play sports and connect - resulted in a project which is unique and new in our country.

So far, 5 seasons of the Quantox IT League and 4 seasons of the Summer League have been played in over 800 matches and with a common goal - good fun, quality games, and an opportunity to meet and connect with colleagues from the industry.

With Luka Bulatović, the director of the IT League, we discussed the idea of ​​gathering IT companies in one place through a sport and the advantages this kind of competition provides and promotes.

‍

The Quantox IT league was formed 6 years ago. What was the initial motive behind the establishment of the corporate basketball league and has the main idea changed to this day?

Six years ago, we gathered at the basketball court where we played basketball once a week. Back then we were teammates on the court, and today, we are the organizers of the Quantox IT League. We moved from the court to the offices and thought about how to connect with other IT professionals and set a new standard that the IT world deserves.

We are aware of the fact that it is very difficult to find free time for physical activity, but when you have a team that motivates you - then basketball training and games become an immense pleasure.

To this day, the same idea remains and we are proud of the entire organizational team as well as all the companies that have trusted us and have been participating for years.

‍

What is the main vision of the Quantox IT League and where do you see it in the coming years?

The idea for this project arose as a desire to present the entire competition in a different light. Our mission is to create an inspiring, creative, and positive environment that encourages unhindered freedom of expression through sports. The club's vision is for people from the IT sector to play sports and meet new colleagues from the profession and also raise their work to an even higher level through the Quantox IT League.

For now, the League gathers only basketball players, but we plan to expand this project to other popular sports. The biggest challenge is to attract more companies that will be part of the league, and we plan to do that with constant improvement of conditions and through introducing novelties.

‍

More than 50 IT companies and about 1500 players have gone through the IT League. What are the key impressions of the competition?

We are proud of the numbers that show us that we worked hard all these years to reach the 6th season and 32 teams. During this competition, some of the teams improved their game thanks to the good organization within the team and a large number of training sessions. 

The reasons for participating in the Quantox IT League differ, from recreation, socializing with colleagues, and strengthening team spirit. Furthermore, one of the main competitive motives is raising the trophy at the end of the competition. And yet, above all these, the most interesting part for the participants is the "third half", when they stay after the game and hang out.

What do the companies that are part of the league see as the biggest benefit?

The Quantox IT League offers companies the opportunity for their people to spend quality time participating in a sport & recreational competition with their team. This way, they improve both their physical and mental health. This is one of the ways for colleagues to get closer to each other, get to know each other better, and spend quality time outside the office together, in a different environment. Further benefits are that, over time, members begin to understand each other better, develop better relationships and have more open communication.

‍

What is the main drive of people who are part of the IT League?

Winning and losing are integral parts of every game. Every team has ups and downs. What drives them are the good results they achieve in matches. Another major motivator is the fans who are there to encourage and cheer for their favorites in the stands!

We believe that the conditions in our league are at the highest level in the country, but we can do even better. The real challenge is to overcome yourself, to do something new. Challenges are what actually drives us!

The IT League has recently gathered 32 basketball teams and organized an humanitarian All-Star event for the treatment of four-year-old Helena Živković from Čačak, who suffers from autism. Hardware and Software teams competed to help little Helena, and collected 126,900 dinars from ticket sales. This is just one of the many ways in which this kind of sports competition contributes to individuals and communities when needed.

The fifth season of the Summer IT League is beginning soon, and registrations start on the 17th of April.

See you there as we continue to strengthen the IT community through team spirit together!

‍

How to connect artificial intelligence and shooting stars?

‍

Well, you can find the answers to these questions from our Dušan Pavlović, Data Scientist in Quantox Technology, who will give a lecture PER DATA AD ASTRA on 23th of March at the Geekstone Meet-Up Event.

Dušan will demonstrate such a connection and talk about how computer vision, deep learning, and research on meteors - shooting stars - overlap. 

One of the topics is how relatively easily you can apply your knowledge of neural networks to solve the problem of automatic detection of meteors in astronomic data made by radio detectors. This problem is one of the most complex ones in this field and our understanding of different problems in the Solar System depends on it.

‍

As someone who studied astrophysics, being an author and the host of the science podcast "Radio Galaxy '' while engaged in the physics of meteors and meteorite astronomy for years, Dušan will show one of the examples of how to apply Data Science to scientific data!

The lecture will be held in Serbian and the entrance is free!

‍

Thursday, March 23, 2023

6:00 - 8.00 PM 

Science and Technology Park

Fruškogorska 1, Novi Sad

‍

Save the date and welcome to the World of Data!

Register at the link

https://www.meetup.com/geekstone/events/292131441/?isFirstPublish=true

‍

The traditional spring CS:GO tournament is upon us!

‍

Serbian IT companies will gather from the 3rd to the 6th of April at the CS:GO Vol.4 charity tournament.

Motivated by the important topic and achieved results from the previous year,

we decided to stay with the same goal - in 2023 the IT community will be playing for Serbian maternity hospitals and better conditions for the youngest and their mothers!

Companies can register up to 3 teams and donate the financial funds to the common fund. As in previous years, the start for one million dinars is provided by Quantox.

After the closing of applications, by public voting and filling out an anonymous survey, we will together decide which maternity hospital we are helping this year!

 We hope that the HIT CS:GO crew will expand this year with new companies and teams that will compete, have a great time and show that the greatest victory is humanity.

‍

Register your team by the 31st of March.

Link for team registration and more information: https://hitturnir.com/

‍

It's time to play even harder!

‍

Meet us on 29 - 30 March

‍

We are looking for an exchange of experience and ideas that improve the digital world and shape our future!

Gathering together Haufe experts and IT companies, partners, and clients, the Haufe X360 Summit takes place at the Kameha Grand Hotel in Bonn from the 29th to the 30th of March.

Lecturers from Haufe and Acumatica - the platform on which Haufe is based- will discuss planned system innovations and software improvements. 

We are looking forward to hearing from great people, meeting other Haufe partners, and sharing experiences and ideas.

‍The Quantox Crew will be there!

• Dušan Milojević / Head of Managed Services
• Thomas Priemel / ERP Consultant
•  Darko Šarenac / Partnership and Business Development Manager

‍

You can meet our people and find out more about our services, advanced digital solutions, and models of cooperation. We are happy to have a chance to network with new potential clients and all new partners ahead!

‍

Summit Agenda and more info can be found at the link

Petar Slović, Chief Innovation Officer @ Quantox Technology, will give a lecture at a workshop organized by DaFED at the Rectorate of the University of Novi Sad on March 1, starting at 6 p.m.

At Peter's workshop "Simplifying Full Stack Serverless Development with tRPC in NextJS" visitors will have the opportunity to learn how to use the power of the T3 stack and tRPC to create scalable full-stack serverless applications.

DaFED is a non-profit organization dedicated to creating educational workshops and networking events for designers, developers, and all tech and innovation enthusiasts.

Discover the benefits of using tRPC in NextJS and learn how tRPC can facilitate the creation of secure and efficient serverless APIs without burdensome infrastructure management.

‍

The workshop is free, and you can find more information, registration, and a link to the live stream at https://dafed.org/

‍

‍

‍

Over the past few years, Data and Artificial Intelligence have become the most trending technology used across all industries. Yes, technology (singular) because Data and AI need to work in synergy to generate the best results. 

By the time the term 'Data' was coined, there was already so much of it that IT professionals and computer scientists quickly realized that sorting out and parsing all the data and, at the same time, analyzing the same data would be impossible for humans. They knew that if anything could help, it would be AI algorithms that would be able to derive valuable real-time insights from that much data. 

The merge of these two disciplines created a technological boom changing businesses left and right. But how exactly?

AI for Business

‍

‍

Unlike a decade ago, the internet is now filled with data providing important information about consumer habits, likes, dislikes, personal preferences, and activities. This type of data, when paired with AI technology, can help businesses:

• boost efficiency with various process automation solutions
• improve consistency and speed of service
• use customer insight and review to inform decision-makers
• discover new opportunities for new products and services
• it can detect fraud threats
• optimize supply chain operations

We could list the benefits of AI for business for days. However, many people would still need clarification, a concrete example, about how to get a taste of some of these benefits and how they can improve their business operations with AI. 

‍

‍

Here at Quantox Technology, we believe in a practical approach and the power of learning. With more than 550 IT experts, our team has some of the best mentors in the industry, ready to convey the usage and possibilities of their products in the simplest way. 

Quantox E-commerce AI Solution Use Case

The Quantox team was introduced to an e-commerce client that needed an AI solution to simplify one part of their working process. Our client had an online furniture store where they sold design furniture and artwork. So naturally, they've set up a webshop with thousands of furniture listings where you can filter your search based on your preferences. For example, you could filter furniture by type (beds, couches, sofas, tables), color, manufacturer, material, and many other specifications.

That meant every product needed a string of descriptive words in their specifications for their search filter to work. For example: If you wanted a black table, you would get a listing of any product that mentions the words "black" and "table." But if you wanted a "Sanne Protin black dining table," if labeled correctly, you would get only one result. 

THE PROBLEM

However, the client's problem was not labeling products correctly, rather than the labeling process being money and time-consuming. At that time, their products were being labeled manually, meaning they needed to pay people to input labels individually. 

OUR SOLUTION 

To fix this problem, the Quantox Data & AI team proposed they create an automatic labeling system that will save them valuable resources. We created an AI system that takes an ad as an input and returns "labels" as an output, meaning product type, color, manufacturer, and material. In terms of business, the result was a significant decrease in the labeling cost and time to obtain the labels, which generally sped up the entire listing processing, had a considerable impact on their business, and positively affected their business efficiency.

Translating Human Problems to Data Problems 

1. ASKING THE RIGHT QUESTIONS

So, we were building this system. It's a simple task for a human but could be difficult for a computer. To successfully transfer this business problem into a data problem, we must ask the right questions.

 What would a human do if they had the task of labeling the products?

The trick is to answer these questions in as much detail as possible. For example, some would say, "well, a human would read the description," but the correct answer is, "the human would look at the listing." That means our internal conversation would look something like this:

‍

‍

Through questions, we slowly realize that humans would look at the image and type in what they see. For example, they see a black table, so they would type that it's a table and it's black. They could also read the description to see who the manufacturer is and what material was used. 

2. DRAWING THE RIGHT CONCLUSIONS

The answers to our questions told us our system needs to be able to emulate human behavior. Mainly it needs to be able to "see" the images and recognize objects on them, but also recognize text, read it, and understand the meaning of it too in order to extract the vital information needed for labeling.

Well, what can "identify" objects in AI? Computer vision certainly can. 

And how can a computer read and understand text just like a human? With Natural Language Processing, of course.

‍

3. THE DATA AND AI MODELS

As always, we needed data to apply to our AI algorithm, so we took around 20 thousand pre-existing labels to train our AI models. The diagram below shows how data trained our models and how it was deployed, monitored, and maintained.

‍

‍

On the left side, you'll notice a very important box saying, "hold out a fraction aside." This means we took a small fraction of data from the 20 thousand label batch (about 10-20%), which is the data we did not show to our model. The reason is simple - we wanted to test our model on that data after it's trained. This way, we can see how our model performs on data it hasn't seen before. 

In the middle, you can see that we have two separate models. One that works with images and one that works with text. Each model is tasked to learn different things, and finally, that information is combined to output the desired information. 

The task is simple: "Build a system that says: This listing is selling a lamp, and the color of the lamp is beige-brown-yellow-ish."

We'll learn how this simple task can quickly become a bit more complicated for a computer.

The listings are given as "image+text" pairs, and we're free to choose from a wide arsenal of AI techniques as long as the solution delivers the desired information that is correct for the majority of the time. In the following sections, we describe some methods we opted for.

Our dataset contains around 20 thousand examples similar to this one, containing images of various furniture types: couches, sofas, tables, wardrobes, etc. In addition, the data includes information about where the object is on the image and the product being sold. We used 15,000 examples to train our AI algorithms (training data), while the other 5,000 we used as test data.

Here's an example of what a typical listing looked like.

‍

‍

‍Disclaimer: The product shown in the photo is for illustration purposes only and does not represent the actual product of our clients. It is used solely to explain and demonstrate the capabilities of our AI models.

‍

QUANTOX COMPUTER VISION MODEL

Our computer vision model was supposed to extract information from the image. We expected it to localize a lamp and tell us that it is, in fact, a lamp. To do so, our algorithm needed to perform two tasks: object localization and image classification. 

A bounding box consists of 4 numbers representing the coordinates of the box itself:

X_upper_left, Y_upper_left, X_lower_right, Y_lower_right.

Our model, in essence, is a model that tries to predict these four coordinates in an image and assign the label "lamp" to this listing. It does it in such a way that the predicted coordinates should mostly overlap with coordinates provided by humans. At the same time, the product label ("lamp" in this case) should match the human annotations in most cases. 

‍

‍GOAL: Maximize the overlap and assign the correct label.
Now that we know what the model should do let's see how it does it. 

WHAT ARE CONVOLUTIONAL NEURAL NETWORKS?

Each image was run through a convolutional neural network (CNN) tasked with learning the features of the image. Neural networks are made of neurons. Neurons group into layers. Layers hold numbers used to process our data (by multiplication and addition). Those numbers represent the knowledge of our network. Intelligence even, if you will.

The CNN model we created contains 53 convolutional layers. Each layer transforms the image by applying a convolution operation on the image. Convolution is actually a whole lot of additions and multiplication following the pixel grid of the image and some additional rules.

‍

‍

What that means is, at a high level, we plug an original image into the CNN, and it goes through each of these layers sequentially. The output of each layer is a slightly modified version of the original image. These modifications can highlight contours and shapes of interest in the original image or blur and make insignificant contours and shapes disappear.

We program the network to tweak the convolution operation and figure out what these shapes are on its own. Then, we let the network try multiple times until the results are satisfactory according to a predetermined criterion. This process of tweaking the layers (or, more technically, tweaking the parameters of the convolutions) is known as training the neural network. 

When the image passes all 53 convolutional layers, we add another two layers, which we construct in such a way that we can read what the network thinks the bounding box is and what is the object residing in that bounding box. Finally, we read predictions from these final layers. They're tweaked similarly to convolutional layers during training. When this lamp was successfully detected, we cropped out the predicted bounding box, extracted the dominant colors, and gave it its first label - a "lamp."

QUANTOX NATURAL LANGUAGE MODEL

Now that we know how the computer vision part of the solution works, the question arises: "Why do we need text at all? Isn't computer vision all we need because we got the colors and the product?"

Well, not exactly. What if the lamp was labeled differently, for example, "torch"? In this example, it isn't, but there could be a lot of other instances where labels can be nothing more than misleading synonyms. We can't simply rely on one-word descriptions of the product coming from some people we don't even know who labeled these listings. They might have labeled it for a different purpose, who knows. That's why we're complementing the information we extracted from images with the data from text. We're shrinking the margin of error and trying to output results that align with the industry's terminology as much as we can. 

Plus, the computer vision model can detect a lamp and might even detect some sort of cat on the bottom of the lamp. Of course, it's essentially wrong detection (because it's not a cat, but a llama), but the text helps us deal with these cases as well, and now we'll see how.

So, let's get technical. Now suppose that our computer vision model detected both a lamp and a cat.

‍

‍

Now we have two detections: a lamp and a cat. Certainly, a cat is not for sale, nor is it a real cat. It's a cat-shaped pedestal for the lamp, which is probably not even a cat but a llama. This is just to give you a sense of ways these models can be wrong. We need to filter the object detection algorithms' outputs by analyzing text. Remember that the listing description states:

"Created and designed by the famous 20th-century art deco designer Jean-Michelle Jarre, this nightlight can be a great addition to your living room or bedroom for only $99.99". 

Our goal is to search for the words "lamp" and "cat" in this text and find a match. But we would fail if we want to match the exact string of words. There are no words "lamp" or "cat" in the description. But there is the word "nightlight." You'd agree that refers to the lamp. Of course, you would, but how do we build an AI system to figure that out?

That's the goal here. And the tool is, again, a neural network. Specifically, the neural network performing the task of "word embedding," unlike the former network performing the task of "object detection." 

The neural network's inputs are text sequences, and the final output is a numerical representation of the input word. Literally, a sequence of numbers. How many numbers? We can decide that, but the common practice is higher degrees of 2, such as 256 or 512. 

So yes, sequences of numbers, like the lottery, but with one crucial property. When two words of similar contexts, such as "nightlight" and "lamp," are represented numerically, they have similar representation, again by some predefined criteria. However, when words of vastly different contexts are represented numerically, these representations are different, such as "lamp" and "truck," for example.  

So we create a neural network consisting of layers, except they're not convolutional. They're different types of layers, each connected and the final layer from where we can read predictions. Like most NNs, this one also has to predict something, and by tweaking this network, it learns which words frequently co-occur, effectively meaning that the network learns which words are contextually similar. Its knowledge is represented in its layers, which we then extract from the network and use as our "numerical representation of the words themselves." 

When this is trained, we get a numerical representation of all the words from the listing description and all the words from the object detection model, namely, "lamp" and "cat." Now we measure the similarity between the word "lamp" and all the words from the description individually and then repeat the process for the word "cat." The similarity score is a cosine (remember trigonometry?) between the numerical representations of two words for which we calculate the similarity. This cosine similarity ranges between 0 and 1, 0 meaning not similar, and 1 meaning identical. 

‍

‍

Now relying on the fact that the listing descriptions are highly likely to mention the product itself or at least some products' synonyms, we can expect that whatever the product is ("lamp" or "cat"), it will have a high similarity score with whatever the product is in the description. 

• In this case, the words "lamp" and "nightlight" have, say, the highest similarity between all the combinations, so we say - "THE PRODUCT IS THE LAMP."  
• The word "cat" will have minimal similarity with the words in the description because the description never mentions animals, cats, dogs, pets, or anything within that context, and context is learned in the neural network, so "cat" will end up omitted. That's how we know - "THE PRODUCT IS NOT THE CAT."

Now let's see the results. 

THE RESULTS

When all this is said and done, and the models are trained and validated, it's time to put them to the test and see how they behave on the data they weren't trained on. Remember that the test data contains the "text + image" and the "location + object" parts. We hide the "location + object" part, obtain the predictions from it, and then compare them with the actual location and object. 

When predicting a product, we achieved 91% accuracy on the training set of around 5k examples. As for colors, the number is at about 70% on the same dataset, but it is a pessimistic estimate based on sampling and employing the naked eye to method to say "this is right" and "this is wrong." The reason for that is that people see colors differently. In the case of our lamp, one person can say "beige," but the other person might say "light yellow." Suppose you convert the color names into an RGB system. In that case, they can differ vastly, so if we calculate the Euclidean distance between "beige" and "light yellow" in the RGB coordinate system, we can wrongly conclude that the model made a wrong prediction. In fact, both the model and the person labeling are right. We believe the actual number is much higher than 70% since the errors mostly come from the edge cases where you can argue it's the same color.

‍THE DEPLOYMENT

Once we've run all data through our models and our models were trained and tested, we deployed the model in production. This means that we need to save the models in some sort of model storage, which could be on some location on our server, on our client's server, or on a cloud storage service of choice. In this case, it was the client's server.

Specifically, we used Flask to create a Python API and Docker for deployment. When the API is called, it triggers a forward pass through the neural networks, which returns some predictions. The application then processes those predictions as described above and returns a response. Finally, the client integrates with our model by calling our API from their application, which gets them the desired information. 

They can then process the information however they see fit, e.g., display them in some sort of drop-down menu on their front end or have radio buttons to help their users filter their search when buying furniture. Here's a diagram describing the solution's architecture at the highest level.

And that's it. We described the process of training a computer vision model, a natural language processing model, we touched neural networks a bit, shining a light on what it means to train the neural network, what are "As" and "Is" in "AI," and how the information obtained from these networks is used in practices to improve the user experience when shopping for furniture online! A lot of math and engineering for such a seemingly simple task, right? However, the work pays off when implemented correctly, especially in the eCommerce industry, where there's a great deal of data and many different types of applications for AI solutions.

‍

Data, AI, and Quantox - a Perfect Recipe for eCommerce Growth!

Transforming your eCommerce business paradigm in 2023 can make your business stand out from the competition by increasing revenue, conversion rates, and the overall popularity of your brand. Integrating AI technology in your eCommerce business gets that much easier with Quantox Technology team of experts that have been providing Data and AI services and eCommerce solutions for years and have a proven track record of success. Our team will take the time to understand your unique needs and develop a customized solution that fits those needs. So, are you ready to unleash the potential of AI for your eCommerce? Let's Talk!

Propel Your Business and Find Your New Team Of Developers @TES Affiliate Conference

‍Lisbon | February 22-25, 2023

‍

Are your software projects delayed due to a lack of developers?
Are you finding it hard to hire and retain good developers?
Are you trying to reduce costs and increase the flexibility to scale your project?
Are you looking for a partner you can trust?

‍

Quantox Technology is coming to Lisbon with 50 developers available right now!

Complete software development from scratch and all-round IT support services.

‍

Pinpoints!

• 16+ years of experience
• 550+ experts in the diverse technology stack
• 13 offices & 7 countries
• 250+ successful projects
• Super-quick scaling
• ISO 9001 and 27001

Boost Your Business!

• Thorough understanding and approach
• Flexibility & adaptability to each client and change
• Managed team & scaling resources to your project
• Efficiency, excellence, and trust

See You @ TES Affiliate Conference!

3 days, 2000+ participants , 75+ speakers, 200+ exhibitors, 80+ countries

The Oitavos Hotel, Cascais/Lisbon, Portugal
Quantox Booth Location: MMT6

‍And Meet Our People!

Schedule a Meeting

Alek

If you are looking for a man who will answer your questions about potential partnership opportunities, look no further. Alek is our CDO with solid skills in leading teams and running projects developed over the years.  Do you have projects that need reinforcement, and engineering support or do you need a partner for brand-new software solutions? Talk to Alek and find out more.

Schedule a Meeting

‍

Dan

He is your guy if you have a major project planned or ongoing but don't know quite yet how to ensure your software development is running smoothly. As our COO, with more than 15 years of experience in the IT industry, Dan can understand you and optimize your software development process.

Schedule a Meeting

‍

‍

Welcome where ideas are made possible!

‍

A new milestone in the IT world! Building the future with strong partnerships!

Quantox Technology and Sandberg Capital have signed a partnership for the further expanding growth of the company, bringing it to a higher level! A €20 million investment is aimed at making a significant leap for Quantox’s future sustainable development, talent acquisition, and strategic global expansion.

Sandberg Capital is a Slovak private equity company established in 2014 with an AUM exceeding €340 million. It focuses on investments in small and medium businesses in the Slovak Republic and the region of Central and Eastern Europe, including among investors institutional ones, such as the European Investment Fund.

This partnership is one of the top 5 investments in the region so far, signed with companies from Serbia.

With Sandberg’s support, Quantox Technology plans to make successful breakthroughs in new markets, building even stronger client relationships with greater freedom and creativity in creating new software solutions.

“By entering into a partnership, we want to broaden our expertise and further strengthen our position as a leading IT employer in the region. Also, this investment will enable us to have a greater presence in the European and US markets, which will mean a lot to our clients by increasing the range of services in delivering high-end digital solutions and providing comprehensive IT support”

Vuk Popović, founder of Quantox Technology

The company’s focus remains the same – discovering new perspectives and empowering the Q team. The partnership is an additional lift that will bring Quantox even closer to long-term growth, sustainability and stability.

“Given Sandberg’s multiple investments in the IT sector, we see an increasingly strong push for digitalization across the economy. At the same time, we perceive a severe shortage of IT professionals that may be preventing companies from growing and achieving their strategic goals. The investment in Quantox reflects our long-term strategy of partnering with ambitious entrepreneurs in sectors that directly or indirectly help with the digital transformation of the economy,”

Michal Rybovič, Partner at Sandberg Capital.

Matej Klenovsky, Investment manager at Sandberg Capital, believes that Quantox and Sandberg share a similar growth mindset and have the same ambitions and perspectives for future achievements.

“Quantox, in addition to employing a large number of IT professionals and having a great ability to attract clients from various industries, is truly committed to the development of the local IT community. We believe that this company is more than ready for the next step and we are eager to follow them along the way, giving support through our experience, knowledge, and capital.“

Matej Klenovsky, Investment manager at Sandberg Capital

Both partners will be focused on strengthening Quantox’s business – establishing a local presence, expanding the spectrum of services to respond to clients’ demands from all parts of the world, and improving internal processes and organizational structure.

“We are extremely grateful for the trust and support of our new partners – this is a kind of confirmation for everything we have done so far, as well as the opportunity to be even better and grow together. Following the shared vision for Quantox, we strive for global expansion and adoption of new knowledge, skills, and experiences while preserving the values ​​and culture of the company itself”

Vuk Popović, founder of Quantox Technology

With 16 years of experience, Quantox is recognized as a reliable partner in the world of digital technologies. After three successful acquisitions in the last year, this is an important step and kind of “wind at our back” that will lead us closer to our vision – to grow in expertise, nurture people as the priority value and become one of the most trusted and efficient partners in the digital world.

Quantox was advised by Grubišič & Partners – Corporate Finance as financial advisor and Four Legal as legal advisor and Sandberg Capital was advised by ESFA as financial advisor, EY as financial & tax due diligence advisors, and BDK Advokati as their legal support.

A new digital era is at our doorstep. By supporting clients worldwide and developing IT potential in our region – we continue making ideas and vision possible!

‍

It’s the final curtain call for 2022 and, as 2023 is slowly approaching, it’s already clear that Data & AI (Artificial Intelligence) technologies will play a major role for many businesses in the future. Companies around the globe have already switched their business model and, in some cases, mindset to the digital world and are now looking for ways to bridge the gaps they found in their digital experience. The best solution is now also the talk of the town- Data & AI technology.

With such a wide range of uses in business, from streamlining job processes and aggregating business data, AI is ready to take the two-dimensional technology and surround the consumers with a three-dimensional physical environment.

We probably already interact with AI on a daily basis without even realizing it. However, it is not the same AI most of us imagined, culturally influenced by decades of science-fiction books and movies. Instead, today’s AI can be something most people, especially younger generations, consider a norm, like ”Alexa”.

The idea of AI is not new, but it certainly took a long time to develop. AI technology came into existence in 1956, and it would take decades of developing the technology and systems to get what we have today. Now, Data & AI go hand-by-hand and are even applicable, soon to become invaluable, to any business if they want to ensure success and steady growth. With the potential to simplify the mundane work process and achieve breathtaking results, AI technologies are becoming necessary if you want to gain a competitive edge in any industry. What was once a theory is now a tangible business toll perpetuated by tech giants like Microsoft, Google, and IBM.

So, if you are a business owner looking for a way to transform your business, this article will help you learn about the benefits of Data & AI in business.

What is Artificial Intelligence (AI)?

In order to fully understand how AI technologies can help a business, it’s essential to know what they are. There are many definitions of ”Artificial Intelligence.” Still, the broad term is that ”AI is any computer software that engages in human-like activities like learning, planning and problem-solving”. A bit vague, huh? It’s like saying your colleague is a ”human”- it’s technically correct but doesn’t specify anything. We must go beyond the basic definition and scramble through the details to learn more about different types of AI and where artificial intelligence is used.

Machine Learning

One of the most used AI technologies in business must be machine learning. Machine learning uses algorithms that process large amounts of data in an extraordinary short amount of time. Plus, those algorithms have the ability to ”learn” over time. The more data the algorithm processes, the more its modeling improves.

For example, you own a factory with all sorts of different machinery, and you need each and every one of those to work like a clock. This is where AI comes in. A large amount of data about the functionality and production of your machinery goes into a central location where, thanks to AI technology, a human worker is presented with real-time information about the condition of your machinery.

Such quantities of data and patterns are impossible for a human to sift through, and that’s why AI technology plays an important part. Machine learning can process all the data more rapidly, recognize the patterns and anomalies, and finally notify the decision-makers if maintenance or a repair is needed.

Deep Learning

The development of an interconnected web of artificial intelligence ”nodes” is also known as deep learning. It is a specific type of machine learning relying on neural networks to engage in nonlinear reasoning. Sounds confusing, right? So let’s try to simplify it.

There’s been a lot of hype around self-driving cars, and understandably so. The idea that a vehicle could make the right decisions in traffic as a human would is revolutionary, but thanks to deep learning, today it’s a reality. Self-driving cars use deep learning algorithms to process data collected from their sensors. Information like the distance from other objects and their speed, predictions of where those objects will be in 10 seconds- all that data helps self-driving cars decide when to change lanes. Moreover, deep learning models can improve with every bit of data they process.

Data & AI in Business Today

While the idea of AI was still fresh and developing, many feared that AI could and would replace humans and their job positions. Still, rather than taking over the world, AI today serves as a supporting tool for all of us.

AI technology still has some difficulties completing common-sense tasks in the real world, but what it can do is process data- and fast! By analyzing data faster, AI can also present us with every possible solution for our problem and the consequences of each action. This can help businesses streamline the decision-making process.

For decades, data collection has been a central business practice but impossible for any individual and company to use efficiently. Why? There’s simply not enough time to interpret all data. However, with AI-enabled applications and other digital tools, businesses can collect real-time data from an established database and learn about their consumers’ problems and preferences. Once they know that, they can easily increase efficiency, personalize their product or service, make informed decisions, and reduce costs. All this will eventually lead to increased revenue.

Common Uses of AI in Business

While there’s more room for AI to grow, as it inevitably will, today, we use artificial intelligence technology in the following ways.

• Machine learning: is often used in systems with a large amount of data. AI helps contextualize that data so the company’s decision-makers can understand energy usage or maintenance demands.
• Customer Relationship Management: CRM systems are rapidly changing with AI technology into a self-updating and auto-correcting system that helps businesses stay on top of their relationship management.
• Internet and Data research: AI has become irreplaceable when it comes to providing internet users with customizable experiences. Now, businesses of all sizes and types can easily target their specific audience without relying on the users’ search box. Thanks to AI, the right information is finding its way to the user now.
• Digital Personal Assistant: businesses can leverage Data & AI in many different ways, not just by transforming their services and customer relations. Artificial intelligence can help reorganize a company’s operations and streamline the workflow. Everything from emails, calendars, and work processes can be done with the help of AI personal assistant bots. So, by leaving smaller tasks to AI, employees, managers, and decision-makers can dedicate themselves to developing strategies that will increase revenue and boost company success.
• Cyber-security: AI serves as a 24/7 ”extra set of eyes” that monitors data input patterns, finds anomalies and threats, recognizes cyber-attacks, and backtracks to their source. It can also prevent future threats.

How Can AI Improve Business Processes?

Do you want to start using artificial intelligence technology in your organization but still don’t know which processes you can and should automate with AI? The latest Data & AI reports show that many industries that implemented this technology are already experiencing major growth in revenue just by automating a portion of their business processes. Well, let’s learn more about the stronger points of AI technology to know how a business can use artificial intelligence to their advantage.

Five Things AI is Just Better At

AI technology is a long way from surpassing human intelligence and has its limits, but there are some things where AI technology is a better solution. Easy to adopt and implement in any business practice, AI is very good at interpreting visual, text, and speech formats.

Visual

Many industries use visual tracking in order to increase productivity and efficiency. This is only possible with AI and its vision-trained capabilities that range from storing and captioning documents to supporting IoT (Internet of Things) sensor arrays.

In short, you can train AI technology to process, identify and search images using computer vision and later use it for things like tracking clinical workflow, automating inspections for a manufacturing process, or even monitoring store conditions.

Speech

So we are back to Alexa or Siri as an ultimate proof of just how much users and consumers love to be engaged. Customer engagement, in this case, excelled thanks to the AI technology converting speech to text and vice versa. You can also use this AI benefit for company staff training.

Knowledge

By mapping information, providing recommendations, and supporting semantic search, AI can be used to improve user experience and offer your consumers information they’ll actually find helpful.

Search

The profile of your everyday user has changed, and newer generations of consumers will be hard to satisfy and engage. Accuracy and search engine speed are the key to success. Still, they can only be achieved by using AI to enhance your search functionality while enabling it to use images, videos, web pages, and more to give your consumers exactly what they need.

Language

We’ve come a long way from communicating to technology in 0’s and 1’s, and today we can even hold a real, full-sentence conversation with it. Thanks to Natural Language Processing (NLP), businesses can integrate bots to chat with consumers. This type of AI technology can translate languages and even recognize sentiment, letting you know whether your consumers are satisfied or not.

How to Manage Data & AI in Business?

With great data comes great responsibility- this is something every business owner should know. You can quickly improve products and services with AI by using the data and increasing its value, but only if you manage and secure your information and protect your business, employees, and customers.

If you own a business that uses AI technology, you should have a comprehensive data management strategy for:

• access
• authentication
• creation
• privacy
• processing
• security
• storage

Cyber-criminals target other businesses’ information by looking for emails, addresses, phone, and account numbers. It is a significant concern of every business, regardless of the size or industry. AI technology can tackle a great deal of tasks related to security of any organization by fine-combining through all patterns and finding any anomaly within milliseconds.

5 Steps to Take Before Implementing Data & AI in Business

Using AI technology for business may be the next best thing, but you should still take steps before transforming your business with AI-based tools.

Step #1 Ask Important Questions

Using the latest AI technology trends can be beneficial for a business only if the right trend is picked and implemented. With that being said, don’t dabble in trends that offer little to no benefits to your business or industry. Instead, ensure that the AI technology is the best tool you can use for your specific industry, business size, and model. Ask yourself if AI can be used to boost your business, but also how you or your users can misuse it.

Step #2 Stay Human-Oriented

Remember that humans created AI technology and, as such, it is prone to mistakes. Dispose of the idea that you’ll create a mistake-free environment by replacing a human employee with AI technology. Always keep this in mind as you review the results your AI provides.

Step #3 Stay Unbiased

Believe it or not, human bias can affect the AI algorithm in many ways, so carefully choose the data set you’ll use as you create a training model. Also, ensure that the creator hasn’t unintentionally programmed their own bias into your AI.

Step #4 Explain Your AI

Your users or stakeholders will want to know the logic behind your AI and how it reached a particular conclusion. Explaining your AI’s decision-making process will help you maintain trust between you and all other parties.

Step #5 Use Data & AI Responsibly

Since AI technology can be quite a powerful tool, it mustn’t fall into the wrong hands. Responsible AI refers to frameworks that ensure businesses act transparently and ethically with AI technology. So ensure you can follow all government regulations regarding using AI technology before you even create an AI business strategy.

The Future of Data & AI

While we can’t predict the future, it is clear that Data & AI technology will become an even more integrated part of the workforce. With the potential to replace many human jobs, this technology can also lead to the specialization of employee skills. Where AI will take over data analysis, people will be required to cultivate qualitative skills like creativity and problem-solving.

With further technology development, we’re bound to see new business applications, consumer uses, and new startups emerging. The Internet of Things and Data & AI make a powerful combination that can easily change the economy. With so many benefits of artificial intelligence, those companies that utilize this technology in the right way can make it to the top in no time, completely turning the tables for themselves and their competition.

Start Your Data & AI Journey with Quantox Technology!

Quantox Technology ensures the best professionals in the field create your AI solutions. Our Data & AI experts have over a decade of experience building reliable AI applications engineered to scale your business and increase revenue. Discover the potential of AI technology and benefits of big data and how it can ensure long-term success for your business with Quantox Technology. Let’s Talk!

‍

Proud to announce that our team has received another reinforcement! We are introducing Marko Nikolić – the New Head of Finance at Quantox.

With many years of experience in various senior positions and expertise in the financial sector, Marko brings a unique set of skills. He is highly creative, innovative, and well-versed in the latest financial trends and strategies.

His passion for finance and enthusiasm for achieving results and high goals make him the perfect person for this position at Quantox. With his expertise in the financial sector, Marko will bring fresh ideas that will help the company develop and grow even more.

What are the expectations, and what is the biggest challenge in working in an international IT company like Quantox?

My primary expectation is to grow professionally together with the company. The biggest challenge of working in a large company like Quantox is coordinating activities with many colleagues from several countries.

In what way will your experience contribute to Quantox in further business?

My experience is complementary to colleagues from the finance sector. I expect that it will contribute to the realization of new initiatives, such as, for example, the implementation of the new budget.

The events of the last few years have led to sudden turbulence in the global market. We see them even today, and they are, to some extent, the cause of changes in how many IT companies work. In your opinion, which strategies of financial adaptation to such changes have given good results, and which could be applied in Quantox – bearing in mind the spread of the company over seven markets?

Quantox has a concrete and straightforward business model, and we should stick to it – “back to basics”!

This can be a competitive advantage in the current global financial trends and can be used in both cases – to expand in the markets where Quantox already operates and to win new ones.

How do digitization and accelerated development of technology affect the role and work of the financial sectors? Do methods and approaches change, and how?

In the Finance and Accounting sector, a sudden jump in the application of Business Intelligence tools for reporting to management and investors was noticeable in the previous years. As a result, CFO-s and their teams had to adapt and improve their reporting. Modern platforms and programs help them in these endeavors; technology is progressing and taking an important place, so constant adjustments are also needed in this sector.

Many factors influence decision-making, especially now. What are the biggest risks, and what are the biggest opportunities you currently see for Quantox’s business?

I believe that the current global situation on the market is simultaneously the biggest risk – it can lead to a decrease in demand for our services – but also the biggest opportunity to expand into new markets under more favorable conditions than before.

What is your main motivation and inspiration in business – what drives you again and again?

I like to innovate, propose and create new things, and constantly look for ways to improve and upgrade that creation.

‍