Case Study: Taranis achieves faster drone image uploads, scalable AI model training, and reduced crop loss with Google Cloud Platform

Taranis Helping farmers to feed the planet with cutting-edge drone imaging and AI

Taranis is a precision agriculture intelligence platform that uses high‑resolution drone, satellite and plane imagery plus AI to help farmers monitor fields, reduce crop loss, and increase yields across more than 20 million managed acres. The company faced the challenge of uploading and processing massive volumes of large images from remote locations with limited connectivity, while needing a scalable, cost‑effective infrastructure to train complex TensorFlow models during seasonal peaks.

Taranis migrated to Google Cloud Platform, using Compute Engine V100 GPUs, Kubernetes Engine, Cloud SQL, Cloud Pub/Sub and Cloud Functions to build an auto‑scaling image pipeline and TensorFlow training environment. The move enabled around 30 TB throughput, processing tens of millions of photos (and ~100 million labeled features), sped uploads from as long as a day to a few hours, cut cost per photo by 10x, improved model stability and release cadence, and made it easier to onboard new customers and expand geographically.

Taranis

Eli Bukchin

Co-founder and CTO