Case Study: Taranis reduces crop loss and scales AI operations with DoiT International and Google Cloud

How Taranis helps farmers feed the planet with TensorFlow

Taranis, a technology company using drones and AI to help farmers reduce crop loss and improve yields, needed a scalable way to upload huge volumes of field images from remote locations and train complex machine learning models efficiently. With DoiT International’s help on Google Cloud Platform, including TensorFlow, Taranis aimed to solve connectivity, speed, and infrastructure scaling challenges as its data and global operations grew.

DoiT International supported Taranis in migrating to Google Cloud Platform, using Compute Engine V100 GPUs, Kubernetes Engine, Cloud SQL, Cloud Functions, and Cloud Pub/Sub to process and serve imagery at scale. The result was faster uploads that went from taking up to a day to just a few hours, automatic scaling for seasonal demand, continuous feature releases, and a tenfold reduction in cost per photo taken.

Taranis

Eli Bukchin

Co-founder and CTO