Frequently Asked Questions

Common questions and responses of AURA Dcap.

faq
    • There are no limitations for the pdfs as long as the pdfs are in the same format in one model.
    • Size of the image should be less than 4mb

  • Yes

  • After the training is done, we can see the average accuracy of the model including the individual %s of the fields.

  • Yes, this falls under the managed service type where we extract the data and perform end- end operation for the customer.

Common questions and responses of AI DRisk.

faq

  • DRisk SaaS has 200+ connectors. DRisk can connect with any ERPs and transactional systems.

  • DRisk has pre-trained models to detect anomalies in online transactions.

  • Pricing is by the transaction. There would be a small one-time cost to create an exclusive, secured tenant to support the use case.

  • DRisk can handle unlabeled data. DRisk unsupervised models are set ensembled AI methods that help to find the needle in a haystack. In some instances, we have seen up to 92% accuracy.

Common questions and responses of AURA Smart Eye.

faq
  • Yes, In fact, SmartEye supports exporting annotations in all standard formats

  • Yes, SmartEye supports Image Classification, Object Detection, and Image Segmentation

  • There are multiple options,

    1. Export in TensorFlow / Pytorch’s standard saved model format
    2. Export as tf-serving docker image with http/grpc endpoints to serve.

  • SmartEye supports three types of data integration

    1. Direct data upload from local
    2. Cloud/Edge storage path mount for large datasets.
    3. Using public URL of video/Images
    2nd option is helpful with large media files.