Tallinn is a toolkit and methodology for machine learning on the Apache Spark open source framework, enabling anyone to quickly become proficient in machine learning and predictive analytics on a big data scale.

Tallinn comes with the world’s first Data Enrichment Engine designed for machine learning data analysis, automatically improving the accuracy and reliability of predictive models.

By automating several important processes and techniques that can often take a new data scientist a long time to perfect, Tallinn will enable your business to see results faster and act quicker - whilst also freeing up data scientists to focus on other tasks that will help drive the business.

tallinn machine learning as a service

What Is Machine Learning?

Machine Learning is a technique that uses your historical data to make predictions about the future - it’s a great way to extract significant value from your historical data.

Machine learning algorithms can understand and learn the complex trends and patterns in your data in a way that humans cannot, and the resulting predictive models can be applied to almost any industry or function.

Example Use Cases for Machine Learning:

  • Predict out-of-stock products
  • Predict sales based on the weather
  • Predict customers likely to churn
  • Real-time upsell / cross-sell recommendations
  • Predict which employees are likely to leave your company
  • Predict which employees likely to become high performers
  • Predict fraudulent transactions or expenses

Key Features

Data Enrichment

Tallinn is the world’s first Data Enrichment Engine designed for machine learning. It takes your raw data and generates 100s or even 1000s of additional metrics to automatically improve the accuracy and reliability of your machine learning models.

Feature Selection

Tallinn automatically selects for you the data items that should be fed into your machine learning models so you only include the attributes and metrics that actually help your predictions, improving accuracy and saving valuable processing time.

Model Selection

Your enriched data is tested against a wide range of machine learning algorithms to help you choose which algorithms are most appropriate for your business case.

Model Tuning

Your machine learning algorithms are automatically tuned to ensure that you are using the optimum settings to maximise accuracy.


Tallinn delivers an end-to-end machine learning process. It provides all of the components needed to deploy, build and run your predictive models in both development and live environments.



Tallinn is based on the Apache Spark framework, it needs nothing else. It can work with almost any data source (via JDBC) or flat file, with no reliance on any commercial 3rd party technology.


Tallinn benefits from Apache Spark’s highly efficient in-memory processing. All work is done offline, so there is zero impact on your source databases.


If you add more CPUs, or even another node to your Apache Spark cluster then Tallinn runs even faster, without any code changes required. You can easily flex your capacity up or down based on your processing needs.

Machine Learning Fast-Track Offering

Peak Indicators are pleased to offer a Machine Learning Fast-Track service to enable organisations to rapidly see the benefits of machine learning and have their staff trained to deliver high-value predictive analytics:

3-day on-site training: “Machine Learning and Predictive Analytics”

Proof-of-concept machine learning exercise with one round of Tallinn data enrichment.

Workshop to present the proof-of-concept results and discuss next steps

1-year subscription to use Tallinn

Interested in Tallinn? Contact Us For More Information