Adtelligence
­­Technology
State of the Art
SOFTWARE INNOVATION IN ACTION FOR YOUR CUSTOMERS

Use your data with the help of Artificial Intelligence to increase your sales performance

formel hoch 1

INNOVATION ON ALL LEVELS

AI Layer

The Machine Learning Center enables the use of algorithms without programming knowledge

Data Layer

Adtelligence refines your data and makes it easy & fast to use

Application Layer

Adtelligence adapts to any IT infrastructure through flexible interfaces and open standards & is easy to combine with your existing tools

Infrastructure Layer

SaaS in a Private Cloud Environment – Secure and Flexible

AI & MACHINE LEARNING IN SALES

Enormous amounts of data are collected in online sales and marketing, revealing a great deal about user behavior and preferences. Adtelligence’s main goal is to make this data usable. For this purpose, we use various machine learning algorithms ranging from Markov models to neural networks.

Today, AI, Big Data And Machine Learning are mainly used by IT departments, Business Intelligence and Data Science. Our focus is to make complex technology usable for everyone. Through AI-based data analysis and personalized customer targeting, we support you in your sales activities and help you increase your closings and sales.

To do this, our software analyzes your customer and visitor data, segments target groups, automates the process of addressing them, and optimizes your channels based on all click and transaction data using Machine Learning. Our algorithms are flexible and versatile. We help you optimize them for your business case.

Tickle out the last 5% with neural networks

The Formula 1 of conversion optimization – using neural networks to tease out the last 5% of converison.
Have you long since left A/B testing behind? Automated traffic allocation is kindergarten? Then how about neural networks? Used correctly, you can get even the last 5 percent out of the converison rate. This is the Formula 1 car for your conversion increase: difficult to drive, but extremely performant.

Contextual Bayesian Bandit

Personalize content based on data

Not every website design or email is equally successful with every customer. We optimize the targeting of users by showing each visitor a variant tailored to them. Adtelligence offers several options for optimizing this playout: Thus, both a neural network and a Bayesian Bandit can be used to optimize on a parameter basis.

In common A/B test scenarios of similar tools on the market, different variants are tested against each other and a preferred variant is chosen in case of statistically significant differences. Our machine learning algorithms work iteratively and converge much faster against a best possible distribution of traffic.

personalisierung

How does it work? (Using the Bayesian Bandit as an example)

The underlying mathematical problem is a so-called “multi-armed bandit” problem. Here, there are several triggers (~variants) with different unknown odds (~conversion rates), which are to be drawn with the largest possible total profit.

One of our solutions to this is a further development of the Contextual Bayesian Bandit. This calculates the probability densities of the real conversion rates of different variants based on previously collected data. A conversion rate is then drawn for each incoming user in these probability densities and the best one is played out.

What’s the benefit?

This algorithm converges much faster towards an optimal traffic distribution which also offers the possibility to optimize for a context of users, i.e. individual probability densities are estimated for each context parameter. The algorithm then recognizes independently and automatically that, for example, smartphone users from northern Germany prefer a different website variant on Mondays than desktop users do on Fridays, and plays out the optimized design, product or approach to the respective visitor accordingly.

Trends and temporal developments in user preferences can also be detected by dynamically weighting the data in the densities and used for optimization. Our Contextual Bayesian Bandit is already in use at many companies and achieves optimization effects of between 10% and 30%.

Clustering with HDB Scan

To ensure that your messages reach the right target group, we apply various automated clustering methods to your customer data. Using a density-based method based on HDBSCAN and fuzzy clustering, we generate meaningful and usable customer segments from your customer data. All these customer segments can be made directly usable for personalization by our system and provide insights about your customer base.

salesintelligence visualisieren

Process automation

Intelligent process optimization with big data saves our customers time and money. Automatic optimization of business processes, image and text recognition, churn detection and sentiment analysis – there are numerous application fields for Machine Learning and we understand how to train Artificial Intelligence. Towards Hyperautomation – With our software, you can both launch rule-based automated mailing campaigns and have them automatically optimized by the Bayesian Bandit. The Bayesian Bandit learns independently and continuously which email variant and which content achieves the highest conversion rate with which customers.

salesintelligence kampagnen

Our focus is on making complex technology usable for everyone

No more programming, only configuration

We support you in your sales activities and help you increase your closings and sales through AI-based data analysis and personalized customer targeting. Our software analyzes your customer and visitor data, segments target groups, automates the process of addressing and optimizes your channels based on all click and transaction data using machine learning.

Our algorithms are flexible and versatile. We help you optimize them for your business case.

BIG DATA TECHNOLOGIES

EXPERIENCE IS THE BEST WAY TO LEARN. AND IT IS IN YOUR DATA.

Vividly speaking, machine learning algorithms learn from past data to estimate future events. You want to know what the website should look like so users are most likely to convert? Or which email should when be send to whom to make the biggest impact? With machine learning algorithms, your experience and our expertise in digital sales, we can achieve sustainable revenue gains – across multiple use cases throughout the lifecycle of your customers.

The prerequisite for this is a clean database. Adtelligence software automatically checks the quality of your data, displays it clearly and gives you direct recommendations for action to improve data quality.

During each data integration, the data is additionally cleaned automatically and made usable for your specific machine learning use cases.

BIG DATA DATABASES

Intelligent process optimization with big data saves our customers time and money. Automatic optimization of business processes, image and text recognition, churn detection and sentiment analysis – there are numerous application fields for Machine Learning and we understand how to train Artificial Intelligence. Towards Hyperautomation – With our software, you can both launch rule-based automated mailing campaigns and have them automatically optimized by the Bayesian Bandit. The Bayesian Bandit learns independently and continuously which email variant and which content achieves the highest conversion rate with which customers.

PREDICTIVE ANALYTICS MODEL

Your models will not get lost

Over many years, you have certainly already trained and further developed a large number of models. These are not lost with Adtelligence, but can be easily integrated and are a perfect addition to your data.

SIMPLE DATA INTEGRATION WITH REST APIs

The basis for being able to learn from data and use Machine Learning successfully is a clean database. The Adtelligence software automatically checks the quality of your data, displays it clearly and gives you direct recommendations for action to improve the data quality.

During each data integration, the data is additionally cleaned automatically and made usable for your specific Machine Learning use cases.

Data & Privacy

DSGVO is our „home turf“

All data is pseudo-anonymized or completely anonymized before processing in order to remove any personal reference. Personal information is generally not necessary for the analysis and use of your customer data, as it has no significance for the behavior of a user. The following applies to the entire process: All data always remains in your possession.

APPLICATION

Open interfaces and standard programming languages for sustainable maintenance reliability

The Adtelligence platform adapts to any IT infrastructure through flexible interfaces and open standards and can be easily combined with your existing tools.

Techstack/Tools

Application and data

Dev ops

Business tools

Agile programming

To fully leverage the skills and experience of our employees in your projects, we use proven agile processes: The intensive agile collaboration of our cross-functional development teams leads to a high degree of coordination, enables more flexibility overall and shortens the time-to-market. As an active team member, the customer itself can influence the process at every stage of the cycle.

PRIVATE CLOUD INFRASTRUCUTRE

Software as a Service – SaaS solutions scalable without limits

Our services are provided to our customers in our cloud environment. Deployment and update are automated and allow us to provide fast and mostly uninterrupted updates. Our service goes through a series of software and security tests before being made available on production systems.

The cloud infrastructure is operated in a high-availability cluster across multiple fire zones. Here, we rely on branded hardware with high resilience and redundant network and power connections.

PrivateCloud

All Adtelligence services are operated in our own private cloud environment in German data centers. Modern virtualization technologies ensure high availability of our services.

On Premise Solutions

As an alternative to hosting in the Adtelligence Private Cloud, Adtelligence also offers on-premise hosting. The platform is connected to our data center via a secure connection (VPN), which can be used to import updates or make adjustments.

All software components are distributed to a central master server via our data center so that the latest version is available at all times and security-relevant patches can be rolled out. Updates can be automatically installed on the servers during a previously agreed maintenance window.

The installation is carried out on site by a technician from Adtelligence and the customer. After successfully checking access via the VPN, all further settings an updates can be made remotely.

iStock 943066192 scaled

Data and IT security

Your data is stored by our distributed storage system to ensure it is always available. A daily backup is available at multiple locations in the unlikely event that you need to restore it in the event of a failure. Data processing is done in compliance with the DSGVO.

Our IT infrastructure is located in a secured data center and consists of a multi-level firewall concept with hardware and software firewalls as well as a central intrusion detection system (IPS). Daily automatic security checks ensure that vulnerabilities are detected and closed more quickly. Our multi-member IT team monitors all services 24 hours a day 365 days a year.