A global fashion brand were experiencing low purchase and click-through rates from their member email marketing efforts and needed help to improve the user experience.

Using AI to improve the segmentation of customers and products, we were able to present more relevant offers and recommendations to the right customer, leading to significantly greater click-through rates and purchases.

Our client - a large company within the pharmaceutical industry - wished to increase throughput on one of their production lines. This specific line detailed a highly branched process that handled multiple different products, each with their own specific set of production rules. As such, it could not be efficiently modeled using standard linear scheduling software.

Using AI, mathematical optimization and constraint programming, our team modelled the production line including all product specific constraints. The developed optimization engine was paired with bespoke front- and backend solutions allowing for fast and flexible process scheduling.

Sweden's largest producer of ecological herbs and lettuces - wanted to introduce AI into the process for growing their best selling crop: basil.

Using computer vision technology, our team delivered a complete hardware and AI-enabled software solution. Based on deep learning models we used real-time picture data to provide short and long-term harvest forecasts, climate recommendations, and warnings about crop quality.

We helped a rapidly expanding organization to automate the manual CV reading and information extraction process.

With data engineering, document parsing, and feature engineering, we developed NLP models (Swedish & English) for classification and similarity search. We then built a model manager platform, tools and API service containers to integrate the solution with the client’s own systems.

The organization for music copyright and royalties in Sweden wanted better tools to predict the commercial value of a new product.

We helped them build an AI solution that leverages segmentation plus prediction based on historical data for related products in order to provide an estimated pre-launch value of a new product.

An influencer marketing platform, needed help harmonizing content from multiple vendors in order to make consistent recommendations to influencers.

Using AI, we could automatically extract data such as product type and product description from images and integrate it with vendor specific category descriptions to build a recommendation system connecting the right products and influencers.

A new venture aiming to help companies navigate the ever growing flow of information came to us for help building an AI system that would help them to realize their vision.

We trained an AI model that could sieve through large amounts of information, rank it according to relevance, and then provide recommendations. Finally, the model would summarize the relevant information for each company in an easy to understand manner and deliver it to the customer.

An analytics tool for M&A and private equity wanted to improve their customer experience by offering better search and keyword functions.

Using web scraped descriptions of approximately 7 million companies from 10 European countries, we built multi-lingual NLP models to harmonize the descriptions cross language use and home country. ‍Further, we used NLP-based methods for similarity search, classification and keyword extraction and integrated the data on the client’s web site via Docker-based service APIs.

During the COVID-19 pandemic, we worked with multiple stakeholders who were in need of continuously updated analysis and results.

We gathered, parsed and harmonized large sets of data (daily updated) from the national health registry and DNA sequencing to extract and classify mutations and variants of the COVID-19 virus. The results were then visualized in a dashboard containing key insights and stats that all stakeholders involved could view and easily understand.

In an R&D and clinical trial setting, we assisted in the identification and prioritization of cancer neoantigens for targeted immunotherapy.

We developed an AI-driven solution that seamlessly integrates DNA and RNA sequencing via a robust NGS variant calling pipeline. Using machine learning to prioritize neoantigens for cancer therapies, we created a scalable and interactive hybrid cloud data analysis and visualization solution.