SuperAnnotate

Technical Setup & Integration

Python SDK

SuperAnnotate provides a Python SDK for programmatic control of annotation projects:

import superannotate as sa

# Initialize client
sa.init(token="<YOUR_API_TOKEN>")

# Upload items to a project
sa.upload_images_from_folder_to_project(
    project="my_project",
    folder_path="./images"
)

# Export annotations
sa.export_annotation(
    project="my_project",
    folder_path="./output"
)

Key Integration Points

• Data Sources: Connect directly to AWS S3, GCP, Snowflake, and Databricks for seamless data ingestion and export.
• Custom Annotation Forms: Build annotation UIs programmatically or via the platform's visual editor to capture any data modality.
• Workflow Automation: Define multi-stage annotation and review pipelines via API, supporting QA cycles, consensus scoring, and model-in-the-loop steps.
• RLHF & Evaluation Pipelines: Pipe human preference data, prompt-response comparisons, and agent traces directly into fine-tuning and evaluation workflows.
• Security: Supports SSO (SAML/OIDC), 2FA, role-based access control, and enterprise audit logging. Certified SOC 2 Type II and ISO/IEC 27001:2022.