Compatibility

Will Diffgram work with my system?

Levels of Integration

Userscripts Overview
Setting Up Webhooks
SDK
Connections Introduction

Compatible with your Hardware and Frameworks.

Compatible with your hardware from bare metal to AWS, Google, Azure, or Virtual Private Cloud (VPC). Use frameworks like Tensorflow, Pytorch or commercial AutoML services. Connect your models for Pre Labeling.

Customer Interface

Build a Custom Interface

Your team can use Diffgram UI primitives and Diffgram backend services to easily create custom user facing interfaces. For example to deeply integrate and embed AI/ML supervision inside your teams application. Or to customize the interface more to your needs, etc.

Use our SDK

Our python SDK provides easier wrappers for streaming data, managing workflow, and much more.

Annotation Patterns

  • In house
    Known subject matter experts, contractors, or employees
  • Out source
    Bring your outsourcing partner
  • Hybrid

Data Sources Input/Output

  • Your Private Cloud
    With Private Deploy Diffgram has zero knowledge of data
    Train models with your own code

  • Google (GCP)
    AutoML
    AI Hub / ML Engine
    Vision AI, Video AI (Custom training)
    Cloud storage
    Train models with your own code

  • Amazon (AWS)
    Sagemaker
    S3 Storage
    Train models with your own code
    Microsoft Azure

Machine Learning Frameworks

  • Tensorflow
  • PyTorch
  • Custom

Performance Improvement Patterns


Did this page help you?