References - TheBlue.ai

References. Our experience makes a difference.

AI workshops for The European Commission in Brussels

What?

Conducting explanatory workshops for politicians, educating on the basics of Artificial Intelligence, the technologies involved, and possible use cases. Taking part in discussions about the influence of EU regulations on the development of AI in Europe, possible environmental impacts and AI business models.

Why?

The European Commission plans to invest more in AI in the few next years. These Workshops aim at helping politicians understand the technical backdrop and practical applications of Artificial Intelligence.

Text analysis in pharmaceutical industry

What?

Multiple projects containing: Simple yet powerful tool for adverse event detection in social media data, that allows a company to quickly react to possible reputation damage. Social media analytics for new drug introduction. Personal information deidentification and Information extraction in medical texts.

Why?

To automatically find new adverse events published on the internet about a drug, detect who is talking about a drug, and the the context of the conversation, enable analysts to work on data without additional permissions. To track the reactions of customers to a newly introduced drug. To extract the most important information from medical publications.

Technology stack

Python, Hadoop, PySpark, Machine Learning algorithms, Natural Language Processing, social media analytics

AI platform for political consultancy company

What?

Artificial Intelligence driven analytical platform that provides a central point of information for political stakeholders.

Why?

To provide relevant information, track trends, sentiments and dependencies. Provide time and monetary savings as a result of moving away from manual analysis. Give completely new insights from the political data.

Technology stack

Webscrapping, NoSQL, Natural Language Processing, Machine Learning, Python, Django, Javascript, Angular 6, social media analytics

Automated text summary solution for call center company

What?

A tool that changes voice to text and creates summary out of most important and relevant information provided by the calling customer.

Why?

It saves the time of consultants that had to write notes during or after the call manually. Gives certainty that the summary has been made after every single call. Enables extended scenarios for analyzing and optimizing internal processes and customer problems.

Technology stack

Python, Speech-to-text, Natural Language Processing, Machine Learning

Social media analytics tool for events organization – with Re-Work

What?

A tool to analyze social media content related to a specified topic. It Includes built-in dashboards to visualize data geographically and determine any relevant trends.

Why?

To track and score opinions about products or services in real time, see main influencers and react to positive and negative feedback in the social media instantly. To respond to a crisis before it even begins.

Technology stack

Python, Scala, Elasticsearch, Kibana, AngularJS, Bootstrap, Docker

Position estimation for machine construction company

What?

Precise positioning system based on data produced by low energy sensors

Why?

To precisely determine the position of a machine or a vehicle without the need of installing additional devices.

Technology stack

Python, C, MySQL, Machine Learning algorithms, real time processing of sensor data

Predictive Maintenance in car manufacturing company

What?

Consulting and preparation of big data and machine learning architecture for detecting anomalies in data generated by sensors.

Why?

For early detection and prediction of machine damage and break-downs allowing preventative maintenance and counteraction. Thus lowering costs and risks connected with machine failures.

Technology stack

Python, Scala, Elasticsearch, Kibana, AngularJS, Bootstrap, Docker

Epilepsy recognition for Alsterdorf evangelical hospital in Hamburg

What?

An algorithm to automatically detect Focal cortical dysplasia (malformations in the brain causing epilepsy) on medical images.

Why?

To facilitate and improve the work of epileptologists, increase diagonisis, and improve treatment of the patients.

Technology stack

Python, neural networks, deep learning, image processing techniques

Body parts recognition in healthcare with apoQlar

What?

A solution that allows doctors to visualize all anatomical structures of the patient during surgery. Enabling precise placement of models on a patients body, using image segmentation to automatically detect bones, veins and parts of brain.

Why?

To speed up and improve the work of surgeons, allowing them to have all information about a patient on mixed reality glasses, eliminating the need to use traditional MRI images. Thanks to gesture control, the doctor is able to use his hands freely throughout the entire surgery.

Technology stack

Python, neural networks, deep learning, image processing techniques, photogrammetry, unity, C#