We know applied Artificial Intelligence

We can help you apply Machine Learning and Data Science into your business

Why applied AI?

As an industry, we have come a long way concerning AI and Machine Learning. The different components for implementing machine learning are finally here. There is no longer a question of “rocket math” to put them to good use. 

With AI and Machine Learning you can make smarter decisions faster, combine the analysis and prioritization of humans, with the speed of computers, automate processes and use new technologies to solve old problems in brand new ways.

As a company, we firmly believe in software reuse and apply it to everything from ideas to code. Different tools and frameworks are mature enough, and we typically use those to improve or build new solutions for you. Machine Learning is at the threshold of affecting your everyday work. We call this applying AI.

Let us share a real-life example that we did with our customer, Sensenode. They enable their customers to monitor the power consumption of heavy machinery for anomalies. Real-time data is a perfect candidate for machine learning. We produced a compelling proof of concept by employing two students and started to reuse Machine Learning through statistical tools.

Machine Learning is finally starting to become familiar and it is time to understand how to put it to good use in your business.

The Softhouse Way with AI/ML

Many factors play a role when determining whether a Machine Learning project is feasible or not. Softhouse can help you with assessing feasibility and building prototypes, and from our experience, this step is more important than in most other undertakings within IT. If you have a large data set but don’t know how you could use Machine Learning, we can use our “Machine Learning eyes” to look for how we can extract value from the data for you. 

We are passionate about using tools and frameworks with proven track records. We achieve results faster with less bugs and money spent. Equally important is our deep understanding and history of agile ways of working. It is especially beneficial with an iterative approach when applying new technologies such as AI and Machine Learning.

It is a common belief that machine learning and data science do not fit well in agile methodology. Nothing could be more wrong! Creating an MVP and iterating is key to successful Machine Learning projects. In many cases the Machine Learning product itself can be used in the iteration. Also, bi-weekly sprint demos with communication of results are important to align priorities and understanding progress. At Softhouse we have agile thinking in our DNA.

Our Areas of Expertise

Deep Learning

We have many years of experience with Deep Learning, supervised and unsupervised Machine Learning, and Data Science.

Image Analysis

Within Image Analysis and Computer Vision, deep learning opens up a multitude of possibilities. Image classification, object detection and semantic segmentation are powering many recent features all the way from your smartphone to your car.

Anomaly Detection

Anomaly Detection can be done in time series, in images or in tabular data. Finding data points that stand out is crucial for automated monitoring and to transition to Industry 4.0.

Natural Language Processing

Natural Language Processing has seen huge performance boosts by deep learning. Some applications are text classification, sentiment analysis and speech recognition.

Recommender systems

Recommender Systems that tailor content to individuals, can be valuable in almost any system where you have users.

Exploratory Data Analysis

Exploratory Data Analysis, with clustering/segmentation and advanced visualization including dimensionality reduction, can help you understand complex data sets.

Efficient Data Handling

Data is food for AI, and efficient Data Handling is central in any machine learning project. This includes data that is input to the AI, but also data that is produced. It is important to track ML development progress with representative metrics and statistical analyses that reflect real-world performance. It may also be important for you to keep track of bias in performance between different user groups.

Data Centric AI

Data Centric AI revolves around the insight that the input data quality is central for AI performance. Often you get more value by focusing on data standardization and data cleaning, than on tweaking model hyperparameters.

Get in touch with our experts

Kerstin Johnsson

AI and Machine Learning Engineer

+46 73- 912 30 47


Björn Granvik

AI and Machine Learning Manager & Technical Evangelist

+46 705 12 72 68


Want to know more?

We can help you!

    Get in touch with us and we will help you!