DeepHawk

Non-quality in manufacturing costs 3,000 billion dollars to manufacturers worldwide (5 and 10% of their revenue).  Visual quality control by human operators is slow, costly and non-reliable. If done by non-AI-based solutions it is costly and not flexible. As for AI solutions, they are either slow for inference (high latency) if run on the cloud, slow for training if based on deep learning and carrying huge datasets, or too depending on specific dedicated hardwares, therefore not flexible.
DeepHawk has developed a frugal AI and delivers the best detection performance to manufacturers:
·       Strong detection capabilities: we can detect anomalies the size of a pixel
·       Fastest AI training: 30 minutes with 50 pictures of OK-products versus over 3 weeks for competitors
·       Flexibility: single/multi view, still/video images, any types of cameras (infra-red, X-ray, microscopic, endoscopic), any types of products (discrete, continuous, bulk and packaging), no need for AI expertise to use it
·       High data protection: because it is deployed 100% on-premise, no data goes to the cloud.
DeepHaw Inc. is based in Los Gatos (CA) and addresses the US market. Our solution is available for live deployments with a special focus on mechanical parts (e.g. automotive), electronic boards and semiconductor testing.
DeepHawk global HQ is located in France.
Ask for information and demo on www.DeepHawk.ai, or contact@deephawk.ai