zabka-logo

zabka-logo-black

Żabka

mbank-logo

mbank-logo-black

Mbank

ikea-white2

ikea-black

Ikea

ccc-logo

ccc-logo-black

carrefour-logo

carrefour-logo-black

Carrefour

newbalance-logo

newbalance-logo-black

New Balance

seat-logo

seat-logo-black

Seat

decathlon-logo

decathlon-logo-black

Decathlon

t-mobile-logo

t-mobile-logo-black

T-Mobile

neonet-logo

neonet-logo-black

Neonet

orange-logo

orange-logo-black

Orange

eobuwie

Eobuwie

<p>Trusted by the most <strong>disruptive</strong> companies&nbsp;</p>

Store unstructured data - without need to decide ahead of time what kind of data you want to store, or how you want to organize it. 

process-automation

Schema-free organization of data

Terrarium DB is a fully managed engine providing the scalability, reliability, security and performance to meet your business data needs at an unbeatable price

webhook-integration

Scalable and accessible architecture

Queries made on billions of records where the time, context, and order of events are essential for business cases: e.g. funnels, segmentations, flow diagrams like Sankey

ai-driven-segmentations

Extremely fast analytics

Easy to use datastore for time series data. It lets you store, analyze and visualize time series of measurements from POS, clickstreams, sensors, machines and other systems in real-time. 

custom-metrics_9d17a77a1a

Time series data

One storage that can do OLAP queries and direct queries for single object data analysis to make decisions based in real-time (<0-500ms)

unified-360

One source of truth

Core features

<p><strong>Heterogeneous</strong> Multi-Modal Data Ingestion&nbsp;</p>

Database with column-oriented & row-oriented storage in one engine.

Characteristic

Initial release

Current release

Primary database model

Relational DBMS, Document store, Time Series DBMS

Secondary database models

Durability

Cloud-based only

Implementation language

Server operating systems

Data scheme

Typing

Secondary indexes

APIs and other access methods

Server-side scripts

Triggers

Partitioning methods

Replication methods

MapReduce

Foreign keys

Optimistic locking, atomic execution of commands blocks and scripts

Transaction concepts

Concurrency

yes, full featured in-memory storage engine with persistence

In-memory capabilities

Key components 

<p>Main features</p>

winners.png

<p><strong>Ready to step up?</strong></p>

logo.svg

<h3><strong>What's the story behind Cleora's origins? How many people worked on it and for how long?</strong></h3><p>Since the inception of the AI team at Synerise, our ambition has been to quickly and easily process giant heterogeneous interactive data. Existing libraries, such as StarSpace, Node2Vec, DeepWalk, or various graph convolutional networks, did not meet our requirements.</p><p>Each of them had a drawback, like very slow performance, impractical limitation of the maximum graph size or unsatisfactory quality of the results. We needed a solution that would allow us to quickly and accurately calculate graph embeddings with millions of vertices and billions of edges to represent user behavior. Several months of waiting for a calculation result was unacceptable for us.</p><p>The first version of Cleora was created at the beginning of 2019 and was implemented in Scala. It was quickly apparent that the tool successfully replaced all existing graph embedding libraries.</p><p>In the next iteration, at the beginning of 2020, in addition to optimizing the algorithm, we decided to get rid of JVM. The entire solution was rewritten from Scala to Rust, thanks to which we have more control over memory and processor consumption, and the speed more than doubled.</p><p>Initially, a team of several people was involved in creating Cleora. Its development gave us additional opportunities to create a number of solutions based on it, including generation of recommendations, scoring, segmentation and various predictions.</p><p>The experience gathered by the entire AI team allowed us to make Cleora what it is today, a universal and reliable "Swiss Army knife" for computing graph embeddings.</p><h3><strong>What is the purpose of this tool and how can Cleora help entrepreneurs? Who should definitely be interested in using it?</strong></h3><p>Cleora is one of the fastest graph embedding algorithms in existence. It is an essential element for systems operating on data in the form of a network of connected nodes. These are recommendation systems, systems which predict the existence of connections between users in social media (e.g like / follow), or even systems predicting the biological functions of protein networks, which allows for the creation of new drugs.</p><p>No wonder then that such algorithms are created by digital giants such as Facebook and Google, creating a number of new solutions each year. However, Cleora has a significant advantage over these algorithms.</p><p>First, it is much faster. Secondly, it does not require specialized hardware (e.g. GPUs for acceleration of calculations) and, in addition, produces high quality embedding vectors. This means that systems (e.g. recommenders) using Cleory may run faster and with greater accuracy.</p><p>Cleora is capable of processing graphs of hundreds of millions of nodes. In social networks, one node usually corresponds to a single user, so Cleora can be used to process datasets on a global scale, at the level of the number of users of the largest social networking sites such as Twitter.</p><p>The release of the software under an open-source license means that from now on, either a company, an individual or a research institution can use Cleora for any purpose. We recommend Cleora when working with large graphs, especially in conditions of limited computing power. The implementation is available on GitHub.</p><h3><strong>Cleora is expected to work 8 times faster than the PyTorch-BigGraph created by Facebook, and Synerise itself has been appreciated by Microsoft. Has the time come for Polish companies to start to make their mark on the international arena?</strong></h3><p>In the scientific sphere, the Synerise team achieved significant success by winning the Rakuten Data Challenge competition at the SIGIR (Special Interest Group on Information Retrieval) conference. The subject of the competition was creating recommendations in e-commerce, and the organizers included Tracy H. King (Adobe), Shervin Malmasi (Amazon), Dietmar Jannach (University of Klagenfurt), Weihua Luo (Alibaba), Surya Kallumadi (The Home Depot).</p><p>The Synerise publication on methods for detecting the most important features of products which determine the user's interest also appeared in the materials of the ICML 2020 conference. A few months later, at the ICONIP 2020 conference, our article was presented&nbsp;describing a model which recommends similar clothes based on a photo gallery from producers and users.</p><h3><strong>What is the plan for further Synerise activities?</strong></h3><p>Our research goal is to enable automatic and very efficient processing of various data sources that are owned by our clients, both in terms of the quality of the results and the calculation time.</p><p>Graph algorithms can process interactive data typically found in banking, telecommunications, and e-commerce ecosystems. However, there are many other types of data such as images, text, sounds and structured data, and any company looking to improve its performance must be able to seamlessly synthesize all of its data into a form that allows easy and instant real-time predictions.</p><p>The business priority for Synerise is now international expansion to Western and Middle Eastern markets.</p><h3><strong>Why is Cleora open-source? What are the benefits of opening the project to the community?</strong></h3><p>Activities of this type bring companies a lot of publicity, especially if the published tool is of high quality, i.e. it is quick, easy to run and comprehensive, and when it is offered under an open license allowing for commercial use. This is the case with Cleora.</p><p>Synerise wants to stimulate knowledge sharing, following the example of digital giants like Google and Facebook who publish some of their solutions. At the same time, we do not see companies on the Polish technology market, even very large ones, that would share their knowledge in an equally open way.</p><p>Of course, Google or Facebook can afford to publish some of their property for free by being a monopoly in their market. However, Cleora is just one of many proprietary technologies being developed at Synerise, so we do not think that its publication will have a negative impact on our company.</p><p>One should also remember that the use of this type of tools is not easy (although we tried to make Cleora as easy to use as possible) and often in practice it will require assistance in the form of consulting, so we also treat it as a potential source of new clients.</p><p>Cleora underpins some parts of our ecosystem and we continuously implement solutions based on it in many companies. Opening the source code has a positive effect on transparency and increases trust in AI solutions, which for many people are still something incomprehensible.</p><p>The transparent approach has many advantages. Scientists employed by our clients may use Cleora to carry out corporate or personal projects to better understand the principle of operation or to validate our performance claims.</p><p>The recruitment aspect should not be underestimated either. High-quality open-source often becomes an element that attracts the most ambitious candidates.</p><p>The advantage of innovation hubs such as Silicon Valley is largely due to the synergy effect - creating a friendly environment for sharing ideas and inspiring each other. We have taken a bold step in this direction. Opening the Cleora code is an important experiment for us, the consequences of which are sure to come back to us through various endeavors. We will observe them with curiosity and draw conclusions. Within days after publication, the first volunteer contributors have already appeared to bring their improvements to Cleora, which makes us very happy.</p><h3><strong>How can open-source software contribute to the development of a project?</strong></h3><p>The publication of the open-source code allows the development of an informal group of users involved in the development of the tool by introducing their own ideas and improvements. In this way, Cleora has a chance to permanently appear in the catalog of graph embedding solutions, as it is constantly developed and updated.</p><p>Of course, this is only possible if the tool is interesting for the community and provides a source of inspiration. We encourage all Rust and Python developers to contribute to our solution, which is and will remain open. Any suggestions for improvements to the project can be submitted via GitHub.&nbsp;</p><h3><strong>Are there any risks related to sharing your projects as free software? Why do so few companies decide to make such a move?&nbsp;</strong></h3><p>A very important goal of many companies is the protection of intellectual property, understood as a strategic resource. From the company's perspective, revealing a key tool’s key code often means losing a competitive edge.</p><p>However, the advantage of Synerise is not based on one unique technology, but on the synergy of many proprietary solutions, concentrated in one ecosystem. Therefore, we believe that the disclosure of a single or even several tools is not a threat to us. The most important reason for such a small number of similar initiatives may be the fact that most companies in our region use technologies created by someone else and profit from implementations.</p><h3><strong>What can we expect in connection with the development of artificial intelligence in business?</strong></h3><p>We can expect a gradual elimination of secondary and non-creative activities, as well as the superhuman possibilities of synthesizing giant data sets. Companies will invest in more and more AI services to improve their operation, i.e. searching for target groups, targeting properly matched advertisements, accelerating processes inside companies, or accelerating communication between technical support departments and customers.</p><p>The artificial intelligence industry is currently at a very early stage, but the time will come to consolidate it in a few years. Then the players with the most universal and comprehensive offer will remain on the market.</p><h3><strong>Which processes in the future could be automated with the help of artificial intelligence? Can we talk about self-operating companies?</strong></h3><p>It depends on the type of industry, but automation and artificial intelligence allow you to achieve better and better results with less and less effort required on the part of human effort.</p><p>Probably in a few years we will be able to expect the first, initially very simple "self-operating" companies.</p><p>Of course, just like on board of a jet, the human pilot will oversee their operation for many years to come, but thus he will be able to focus his attention on truly creative problems and innovation, instead of performing unambitious and repetitive tasks. Fully automatic grocery stores are being tested even today.</p><h3><strong>How is artificial intelligence affecting our lives right now?</strong></h3><p>Machine learning has been around for decades in hedge funds, banks and other parts of the financial sector. For over a dozen years, they have penetrated into ever wider branches of the economy.</p><p>The boom in recommendation systems in e-commerce companies is already responsible for over 30% of its turnover. Without artificial intelligence, companies of this type today have no chance to compete with market leaders.</p><p>The increasing adoption of AI solutions allows us, above all, to automate the most tedious and labor-intensive activities, not only those that have been performed by people so far, but also those that, due to the enormous amount of work, have so far been beyond the reach of humanity.</p><p>We have an opportunity to observe the next gigantic achievements from impressive models generating natural text on any topic, such as GPT-3 (OpenAI), to models simulating protein folding such as AlphaFold (DeepMind) with unprecedented accuracy. Despite their relative conceptual simplicity, since these models do not yet have much in common with "human intelligence," such advances will revolutionize entire fields of science, and soon also industry.</p><h3><strong>And how does AI affect the everyday life of the average person?</strong></h3><p>AI is everywhere today. Our phone collects data about the pages viewed and displays personalized ads. GPS data analysis systems track the routes on which we travel. We have tools for recognizing human speech (eg Siri, Google Assistant), making translations (language translations), as well as advanced image analysis (thanks to this, we have the ability to unlock the phone using a scan of your own face).</p><h3><strong>What does the artificial intelligence market look like today? Is it easy to find specialists?</strong></h3><p>The artificial intelligence market is paradoxical today.</p><p>On the one hand, there are many applicants who are expressing great interest in machine learning and AI. Deep learning in particular, as a newly created area, is of great interest. On the other hand, universities practically ignored this field until recently. For this reason, most specialists with more than 6-8 years of experience are self-taught. There are very few of them and that is why they are in high demand.&nbsp;</p><p>However, not all work related to the field of artificial intelligence requires the participation of a specialist. The vast majority of companies use solutions created and implemented by others, for which even a general knowledge of programming and data analysis is sufficient.</p><h3><strong>What skill would you distinguish as needed to work on artificial intelligence?</strong></h3><p>There are many positions in the work on AI that can be divided into two main groups: research and implementation.</p><p>Research positions require knowledge in the field of mathematics, understanding the internals of machine learning models and being up to date with the latest developments in this rapidly growing field by reading scientific publications, blogs, and reports. In addition, programming skills in languages particularly useful for modeling AI tools are required, where the most common choices are Python or R. Experience in publishing and knowledge in the scientific world is an additional advantage.</p><p>In implementation positions, an important part of the job are skills related to big data processing. In this case, we emphasize knowledge of databases and technologies dedicated specifically to Big Data or streaming data such as Hadoop, Spark, Hive, Kafka, Clickhouse. We attach great importance to the ability to write high-quality code in languages such as Scala, Java, C++ and Python.</p><p>&nbsp;</p><p><i><strong>Interview with Barbara Rychalska</strong></i></p>

cleaora

 From this article, you will get to know about the secret weapon of cleora.ia and how to use it to stay ahead of the competition

Cleora.ai - Swiss Army knife - An essential element of systems operating on data in the form of a network of connected nodes. | Synerise

Cleora.ai - Swiss Army knife - An essential element of systems operating on data in the form of a network of connected nodes.

Open source framework

Collect information in real-time with behavioral data platform.

Measure what really matters in real time.

Engage & empower your customers by rewarding them.

Integrate all marketing channels into one cohesive, insightful platform.

Ultra-fast disk & in-memory column & row store engine.

Apply science to behavioral data. Automatically.

General-purpose model for efficient & scalable learning.

IMG_5646_1.jpg

orange-logo.png

<p>Synerise cooperates with Orange Polska, the market leader in mobile telecommunications, internet and data transmission, by supporting their digital transformation and providing customers with the best experience of using modern technologies. Synerise solutions help Orange acquire and retain satisfied customers, driving an increase in the share of digital channels in overall sales.</p>

<p>Building better customer experience and growth in e-commerce</p>

Case study

ic_event_note_24px.svg

<h3><strong>Before you start</strong>&nbsp;</h3><figure class="image"><img src="/uploads/image_7e2283c8ea.png" alt="Blue squares, charts and client icon "></figure><p>The process should start even before the actual integration goes ahead. Take a closer look at your website and the <a href="https://synerise.com/papers/multi-modal-embedding-recommender">AI solution</a> you will be using, in order to know how to better plan the implementation.&nbsp;</p><h4><strong>1.&nbsp;&nbsp; Analyze products on the website&nbsp;</strong></h4><p>It’s a good practice to analyze product frames on the website. This will help you speed up the whole integration by preparing an appropriate product catalog (the source of e-commerce product stock). Pay attention to information about products that you already have on the website — maybe apart from the title and price there are some special labels like “eco-friendly”; stars with product ratings; special prices for club members, etc. After you define this, you will know what information should be included in the product catalog and so the frame with recommendations will visually fit the website layout.&nbsp;</p><h4><strong>2. &nbsp;Verify the website structure&nbsp;&nbsp;</strong></h4><p>You can also examine the website in terms of places to put recommendations. Areas like the home, product or category pages are obvious but perhaps there is a better space to add recommendations such as in the drop-down menu or on a landing page which describes one of the brands you sell? <strong>Each website should be analyzed individually to find the places where recommendations can be added to enhance the customer experience.</strong> It will also help you to plan the implementation because you will know how many campaigns will need to be prepared.</p><h4><strong>3. Obtain basic knowledge about the </strong><a href="https://help.synerise.com/ai-7-days/"><strong>AI solution&nbsp;</strong></a></h4><p>You will learn about any platform once you start using it, but it’s important to know earlier about the range of services it offers. You need to go deeper by getting to know the requirements of each AI product offered. That will indicate what you will need to prepare on your side, making it easier to better plan the integration. As a result, you will get an idea of what you should focus on to implement campaigns that you are planning, and what should be provided in the short and long term.&nbsp;</p><h3><strong>Integration</strong></h3><figure class="image"><img src="/uploads/image_9063727d97.png" alt="Product attributes like name, price, color, link, image, category, season"></figure><p>Since you already have a concept in mind of what you want to achieve, all you have to do is take action. Start by carefully planning the steps of the integration and providing events and a product catalog.&nbsp;&nbsp;</p><h4><strong>4. Add custom attributes in the product catalog&nbsp;</strong></h4><p>The <a href="https://help.synerise.com/developers/product-feed/">product catalog</a> is the source of information about your products, which should contain basic data like their titles, images, links and prices. But it’s recommended practice to add more attributes, which will provide more details about each product. Thanks to more detail, you will be able to filter your recommendations in a more advanced way. Those attributes can be universal, like ratings, or strictly connected to a specific industry, like seasons for fashion, or energy consumption in the home appliance industry.&nbsp;&nbsp;</p><h4><strong>5. Include unavailable products in the product catalog&nbsp;</strong></h4><p>If your website includes products which are currently unavailable or which can be bought offline only, add them to the product catalog anyway but with the information that they are out of stock. Then such products will not appear in the recommendation’s frame, but you will be able to show recommendations on a product page. In other words, you can recommend similar products which are available, and which may interest your customers and convince them to make a transaction.&nbsp;</p><h4><strong>6. Provide events as soon as possible&nbsp;</strong></h4><p>AI models need data to learn, so don’t wait with <a href="https://help.synerise.com/developers/web/event-tracking/">event integration</a>. Even if you have to enhance the product catalog (see previous point), provide events for better recommendations. The more of them you add, the more accurate the models will be. Another thing which may be helpful is the import of historical transactional events - it will also allow you to get more appropriate results faster.</p><h3><strong>Campaign implementation&nbsp;</strong></h3><figure class="image"><img src="/uploads/image_7c292b6ce4.png" alt="Zero page results showing recommendation on pages where customers don’t view products"></figure><p>Preparation is a crucial part of integration but now it’s time to finally put into practice the data you’ve provided and create campaigns. Let's look at how to do just that.&nbsp;</p><h4><strong>7. Give recommendations at every step of the customer journey&nbsp;</strong></h4><p>To enhance the customer experience, <a href="https://synerise.com/blog/perfect-customer-journey-based-on-ai-recommendations">add recommendations at every step of the customer journey</a>.&nbsp;</p><ul><li>Put <a href="https://synerise.com/blog/bestsellers-vs-personalized-recommendations">personalized recommendations</a> on the home page where customers will be displayed products best suited for them, instead of the same set of products for everybody.</li><li>When customers go to the categories page, show them personalized recommendations tailored to the current category, placed above or below the list of products.</li><li>On the product page, add similar or visually similar recommendations near the main information about the product.</li><li>When a customer adds a product to the cart, add cross-sell recommendations in the pre-cart pop-up to increase the cart value.</li><li>Finally, when customers get to the shopping cart page, add cart recommendations to interest them in additional products.</li></ul><h4><strong>8. Enrich pages where customers don’t view products&nbsp;</strong>&nbsp;</h4><p>You can also help customers by implementing personalized recommendations on pages where they usually do not see any products, like 404 pages, zero search results, empty carts, pages with campaign rules. This gives you more opportunity to attract customer attention with offers specifically tailored to them.</p><h4><strong>9. Go beyond the website&nbsp;</strong></h4><p>So far, we have covered recommendations for most places on the website, but you should go beyond this and use other touchpoints to promote recommended products. Use emails to send personalized offers when the user ends the session, offer complementary products after they buy something, or show similar products based on the last seen item. Send web push messages that link to a landing page with products chosen for the customer. Transfer personalized products to a remarketing platform so that your customers get better suited advertisements.</p><h3><strong>Effectiveness</strong>&nbsp;</h3><figure class="image"><img src="/uploads/image_7efd25046a.png" alt="Violet KPI icon from squares and triangles"></figure><p>Once your recommendation campaigns are up and running, you have to know how they perform. Being able to measure progress towards your goals is key. You should know how and what you want to measure, even in the implementation phase, to be sure you have everything needed beforehand.&nbsp;&nbsp;</p><h4><strong>10. Set your KPI&nbsp;&nbsp;</strong></h4><p>Decide what you want to check. Revenue? Conversion rate? Average order value? CTR? Be specific in your choice of indicators, limit the number of <a href="https://synerise.com/blog/how-to-choose-and-track-right-KPIs-in-marketing">KPIs</a> you choose to focus on and make sure that you collect all the data needed to count them. Set your goal and regularly <a href="https://help.synerise.com/docs/analytics/">check the results</a> so that you will be able to detect any anomaly.&nbsp;</p><h4><strong>11. Use a control group</strong></h4><p>How do you know if your KPIs are good? You should be able to compare them, but compare to what? To a different time period? A lot of conditions can influence results. To isolate one factor (recommendations) and measure it, you should create a <a href="https://synerise.com/blog/control-group-in-marketing-campaigns">control group</a> – randomly selected users who will not get any recommendations. Thanks to doing this, you will be able to compare two groups - people who see a recommendation and those who don’t. Based on that, you can check the effectiveness of AI campaigns.&nbsp;&nbsp;</p><h4><strong>12. Prepare A/B test&nbsp;</strong></h4><p>If you’re not sure which recommendation will be better on the product page - similar or visual recommendations - or you’re wondering if you should put the product frame higher or lower on the category page, you don’t have to guess. Just prepare an <a href="https://synerise.com/blog/ab-tests-your-way-to-read-customers-minds">A/B test</a>. After a selected time period, you can check the results of each campaign, including its CTR, revenue, number of transactions and products bought, conversion rate and more. Knowing which campaigns perform better helps you to make the best business decisions.&nbsp;</p><h3><strong>Cooperate</strong>&nbsp;</h3><p>The tips above will give you a head start but integration is a process, during which you will learn what can be changed, what else you need and what works best for your customers. With your business knowledge and the experience of the integration team who will support you, you will be able to truly elevate the experience of your customers. Remember, that cooperation is the key to success.&nbsp;</p><p>&nbsp;</p>

Shopping basket with AI icon and plus icon which represent AI recommendations in ecommerce

ai-products-recommendations-in-ecommerce

Here’s what you should pay attention to in any AI implementation journey, in order to uncover the most value for your business. Check the best practices!

How to boost AI product recommendations in e-commerce - best practices! | Synerise

How to boost AI product recommendations in e-commerce - best practices!

On our blog

ic_autorenew_24px.svg

ic_android_24px.svg

ic_all_out_24px.svg

ic_accessibility_24px.svg

ic_book_24px.svg

Join community of winning companies

Partnership program

ic_golf_course_24px.svg

ic_pool_24px.svg

ic_smoke_free_24px.svg

Homepage

Product - Terrarium DB

Ultra-fast <u>disk & in-memory column & row store engine</u> designed specifically for <u>real-time</u> heterogeneous data collection & processing, behavioral intelligence, data management & execution of <u>various business scenarios at scale</u>.


terrarium-main-synerise

<p><strong>TerrariumDB</strong></p>

Ultra-fast disk & in-memory column & row store engine designed specifically for real-time heterogeneous data collection & processing, behavioral intelligence, data management & execution of various business scenarios at scale.

TerrariumDB

Heterogeneous Multi-Modal Data Ingestion