SEGES Innovation and twoday kapacity have partnered to revolutionize agriculture using AI. As Denmark’s top agricultural knowledge and innovation center, SEGES Innovation is dedicated to driving sustainability and innovation in farming. Together, they have created a customized AI platform using twoday kapacity’s Best Practice MLOps Framework, catering to farmers’ specific needs. To enhance their achievements, SEGES has adopted an advanced MLOps platform, automating tasks and enabling dynamic analysis, training, and scalable deployment.
Learn more about the challenge, strategy, and results of the case below!
The challenge SEGES was facing
Leading agricultural knowledge and innovation center in Denmark, SEGES Innovation, has one mission: to put farmers and food companies at the forefront of sustainable and innovative agriculture. Failure to spot and manage even the smallest issue could prove catastrophic for both the animals’ welfare and the quality of the produce. The challenge was to find a way to manage and prevent these failures. SEGES had a solution that was partly on-premise, used a lot of different tools and had a large manual code base. This meant that the data science team spent a lot of resources on maintaining and retraining the models, and the path from model development to production was extended and complex
twoday kapacity’s strategy to solve the challenge
What SEGES needed was a solution that would allow them to train and test their models in the easiest, fastest, and most efficient way – the perfect candidate for twoday kapacitys MLOps Framework.
Over three months, Microsoft and twoday kapacity built a clean and lean data science machine able to bring much-needed flexibility and speed to the SEGES data science team.
SEGES’ results
With hundreds of cameras monitoring cows in stables, SEGES can now spot early signs of disease or injuries in up to 90% of cases enabling sustainable farming practices. Powered by Azure Machine Learning the platform is used by the team to perform a wide range of tasks – from predicting crop yield, to automating and optimizing account reporting, and to leading prevention and management of potato blight (important to avoid diseases). With the MLOps framework SEGES has automated their processes bringing the average retraining time down from 6 months to a single day!
The solution has reduced the maintenance costs by more than 95%. On top, the costs of deploying, running, and monitoring machine learning models in production has gone down by more than 80%.