Case Study: Sharper Shape boosts utility inspection ML speed and efficiency with Valohai

Drones and computer vision for utility inspection

Sharper Shape uses drones and computer vision to inspect utility infrastructure and detect hazards like vegetation encroachment, broken insulators, and other dangerous setups. To support its growing machine learning operations, the company needed a more efficient way to manage large volumes of annotated image data and reduce the manual effort involved in labeling and experiment tracking. It worked with Valohai and Labelbox to strengthen its ML workflow.

Using Labelbox for data labeling and Valohai for end-to-end MLOps, Sharper Shape automated much of its training pipeline and moved from manual, fragmented processes to a cohesive system. With Valohai, the team could run experiments on cloud hardware, automate retraining when new labeled data arrived, and reduce infrastructure overhead; meanwhile, Labelbox helped cut labeling costs by as much as 50%. Sharper Shape also reported speeding up model training by over 10X and reducing onboarding time for new data scientists by about 75%.

Sharper Shape

Jaro Uljanovs

Data Science / AI