# DataFramer

- [What is DataFramer?](https://www.dataframer.ai/docs/index.md): The AI Quality & Business Impact Platform. Improve your product's AI quality. Prove its business impact.
- [Connect your traces](https://www.dataframer.ai/docs/connect-traces.md): Pull your AI traces into DataFramer. No traces yet? Connect a new Langfuse or LangSmith project.
- [Architecture & deployment](https://www.dataframer.ai/docs/deployment.md): Deploy DataFramer in our cloud or in your own cloud account.
- [AI usage and credits](https://www.dataframer.ai/docs/ai-usage-and-credits.md): Use DataFramer credits or your own keys
- [User Signals & Journeys](https://www.dataframer.ai/docs/user-signals/overview.md): Connect what your users do to the AI traces behind it
- [Browser SDK](https://www.dataframer.ai/docs/user-signals/browser-sdk.md): Send user events from your frontend with @dataframer/signals
- [Server Instrumentation](https://www.dataframer.ai/docs/user-signals/server-instrumentation.md): Auto-stamp journey ids onto your AI traces with dataframer-journey
- [Collect and Correlate](https://www.dataframer.ai/docs/user-signals/collect-and-correlate.md): How DataFramer receives signals and traces and joins them into one journey
- [Journeys](https://www.dataframer.ai/docs/journeys/overview.md): Browse a user's full timeline and the AI traces inside it
- [Dashboards & Metrics](https://www.dataframer.ai/docs/journeys/dashboards.md): Build custom dashboards over your journeys, users, and signals
- [Discovery & Tracking](https://www.dataframer.ai/docs/findings/discovery-and-tracking.md): Search your AI traces for the patterns you know and the ones you don't, then pin the ones that matter and watch them over time.
- [Root Causes & Remediation](https://www.dataframer.ai/docs/findings/root-causes-and-remediation.md): For every failure you track, get the distinct root causes behind it, each with evidence and a suggested fix, and read them from your coding agent.
- [Human Reviews](https://www.dataframer.ai/docs/reviews/overview.md): Grade AI traces against a shared rubric and turn expert judgment into reusable memory.
- [Rubrics](https://www.dataframer.ai/docs/reviews/rubric-studio.md): Define the dimensions, instructions, and examples that reviewers and judges grade against.
- [Queues & Reviews](https://www.dataframer.ai/docs/reviews/operational-loop.md): Route AI traces to reviewers, grade them against the rubric, and reuse every correction.
- [Review Copilot](https://www.dataframer.ai/docs/reviews/copilot.md): Grade faster with a copilot that knows the trace, the rubrics, past reviews, and the user journey.
- [Calibrated Judges](https://www.dataframer.ai/docs/judges/overview.md): Build an automated judge from your rubric and calibrate it against human review until you can trust it.
- [Dataset evaluations](https://www.dataframer.ai/docs/judges/datasets-and-evaluations.md): Build benchmark datasets of real traces and run a judge across them.
- [API guardrails](https://www.dataframer.ai/docs/judges/guardrails.md): Call a judge from your own code to grade content the moment it is produced, and act on the verdict.
- [Generation quickstart](https://www.dataframer.ai/docs/quickstart.md): Generate your first synthetic dataset in 5 minutes
- [Core Concepts](https://www.dataframer.ai/docs/concepts.md): Understand how DataFramer generates synthetic data
- [Complete Generation Guide](https://www.dataframer.ai/docs/workflow.md): Step-by-step walkthrough of all DataFramer features
- [Data Anonymization](https://www.dataframer.ai/docs/detection-anonymization-augmentation.md): Detect PII, PHI, financial data, identity documents, and more—then redact your datasets using AI models and pattern-based rules
- [Basic Use of Python SDK](https://www.dataframer.ai/docs/tutorials/basic-python-workflow.md): End-to-end example: create a spec from a description, generate samples, and download results.
- [Folder Generation](https://www.dataframer.ai/docs/tutorials/multi-folder-workflow.md): Complete guide to working with multi-folder datasets using the DataFramer SDK
- [PII/PHI Anonymization](https://www.dataframer.ai/docs/tutorials/pii-phi-anonymization.md): Detect and mask Personally Identifiable Information and Protected Health Information in your datasets using the DataFramer SDK
- [Fine-Tuning with Synthetic Training Data](https://www.dataframer.ai/docs/tutorials/finetuning-with-synthetic-data.md): Use DataFramer to generate diverse training data from a handful of examples, then fine-tune an open-source LLM and evaluate style transfer.
- [Financial Bank Statement Extraction with Golden Labels](https://www.dataframer.ai/docs/tutorials/financial-bank-statement-extraction.md): Generate synthetic financial statement PDFs with co-located Q&A golden labels using a multi-folder seed dataset, then evaluate an LLM's financial data extraction accuracy on the generated documents.
- [Support Chatbot Broader Evaluations with Contextual Eval Sets](https://www.dataframer.ai/docs/tutorials/support-chatbot-broader-evals.md): Go from 11 hand-labelled seed rows to a 50-sample (1000s if needed) eval set — complete with golden labels — by using DataFramer to generate targeted edge-case interactions for your support chatbot.
- [Insurance Underwriting Model Drift Detection](https://www.dataframer.ai/docs/tutorials/insurance-underwriting-model-drift-detection.md): Simulate a shifted life-insurance applicant population with DataFramer, score it with a frozen underwriting model, and inspect whether the change looks like true risk or calibration drift.
- [Fraud Detection: Labeled Training Data for Impossible Geographic Jumps](https://www.dataframer.ai/docs/tutorials/fraud-detection-geo-jump.md): Go from a 10-row unlabeled credit card transaction seed to 100s or 1000s of fraud-labeled training samples by teaching DataFramer a concept that never existed in your data.
- [Databricks integration](https://www.dataframer.ai/docs/integrations/databricks.md): Overview of the DataFramer - Databricks integration. DataFramer is a Validated Databricks Partner.
- [Single File](https://www.dataframer.ai/docs/integrations/databricks-single-file.md): Generate high-quality synthetic single-file datasets within Databricks using DataFramer.
- [Multi-File & Multi-Folder](https://www.dataframer.ai/docs/integrations/databricks-multi-file.md): Generate high-quality synthetic multi-file and multi-folder datasets within Databricks using DataFramer.
- [API & MCP](https://www.dataframer.ai/docs/api-and-mcp.md): Two ways to use DataFramer outside the UI: the Python SDK and MCP (Model Context Protocol).
- [List failures](https://www.dataframer.ai/docs/api-reference/failures/list.md): List the failures DataFramer tracks in your production traces
- [Get failure](https://www.dataframer.ai/docs/api-reference/failures/retrieve.md): Retrieve a specific failure by ID
- [Get context](https://www.dataframer.ai/docs/api-reference/failures/context.md): Retrieve the knowledge and rubric content that failures reference
- [List judges](https://www.dataframer.ai/docs/api-reference/judges/list.md): List the workspace's judges with their latest ready version
- [Get judge](https://www.dataframer.ai/docs/api-reference/judges/get.md): Retrieve one judge with its latest ready version
- [Evaluate content with a judge](https://www.dataframer.ai/docs/api-reference/judges/evaluate.md): Grade content against a judge's rubric and get the verdict inline
- [List judge evaluations](https://www.dataframer.ai/docs/api-reference/judges/list-evaluations.md): List this judge's evaluations, newest first
- [Get judge evaluation](https://www.dataframer.ai/docs/api-reference/judges/get-evaluation.md): Retrieve one evaluation with its full verdict
- [Create seed dataset](https://www.dataframer.ai/docs/api-reference/datasets/create.md): Create a new seed dataset with uploaded files
- [Create seed dataset from ZIP](https://www.dataframer.ai/docs/api-reference/datasets/create-from-zip.md): Create a seed dataset by uploading a ZIP file
- [List seed datasets](https://www.dataframer.ai/docs/api-reference/datasets/list.md): Get all seed datasets for the user's company
- [Get seed dataset](https://www.dataframer.ai/docs/api-reference/datasets/get.md): Retrieve a specific seed dataset by ID
- [Download seed file](https://www.dataframer.ai/docs/api-reference/files/download.md): Get a URL for downloading a seed dataset file
- [Delete seed dataset](https://www.dataframer.ai/docs/api-reference/datasets/delete.md): Delete a seed dataset and all its files
- [Create spec](https://www.dataframer.ai/docs/api-reference/specs/create.md): Generate a new spec (async operation)
- [List specs](https://www.dataframer.ai/docs/api-reference/specs/list.md): Retrieve information about all existing specs
- [Get spec](https://www.dataframer.ai/docs/api-reference/specs/get.md): Retrieve a specific spec by ID
- [Update spec](https://www.dataframer.ai/docs/api-reference/specs/update.md): Replace a spec with a new version specified using YAML
- [Delete spec](https://www.dataframer.ai/docs/api-reference/specs/delete.md): Delete a spec by ID
- [Start a new generation run](https://www.dataframer.ai/docs/api-reference/runs/generate.md): Start a new data generation run
- [List runs](https://www.dataframer.ai/docs/api-reference/runs/list.md): Get all runs for the user's company, ordered by creation time (newest first)
- [Get run](https://www.dataframer.ai/docs/api-reference/runs/get.md): Retrieve a specific run by ID
- [Download all generated files](https://www.dataframer.ai/docs/api-reference/runs/download-all.md): Download all generated files from a run as a single ZIP archive
- [Download generated file](https://www.dataframer.ai/docs/api-reference/runs/download-file.md): Get presigned URL for downloading a generated file
- [Cancel run](https://www.dataframer.ai/docs/api-reference/runs/cancel.md): Cancel a running or pending generation job
- [Delete run](https://www.dataframer.ai/docs/api-reference/runs/delete.md): Delete a run and its generated files
- [Create evaluation](https://www.dataframer.ai/docs/api-reference/evaluations/create.md): Start a new evaluation for a completed run
- [List evaluations for a run](https://www.dataframer.ai/docs/api-reference/evaluations/list.md): Retrieve all evaluations for a specific run
- [Get evaluation](https://www.dataframer.ai/docs/api-reference/evaluations/get.md): Retrieve full evaluation details including distribution analysis
- [Create red team spec](https://www.dataframer.ai/docs/api-reference/red-teaming/create-spec.md): Create a new red team specification
- [List red team specs](https://www.dataframer.ai/docs/api-reference/red-teaming/list-specs.md): Get all red team specs
- [Get red team spec](https://www.dataframer.ai/docs/api-reference/red-teaming/get-spec.md): Get a specific red team spec
- [Update red team spec](https://www.dataframer.ai/docs/api-reference/red-teaming/update-spec.md): Update a red team spec
- [Delete red team spec](https://www.dataframer.ai/docs/api-reference/red-teaming/delete-spec.md): Delete a red team spec
- [Create red team run](https://www.dataframer.ai/docs/api-reference/red-teaming/create-run.md): Create a new red team run to generate adversarial prompts
- [Get red team run status](https://www.dataframer.ai/docs/api-reference/red-teaming/run-status.md): Get the current status of a red team run
- [List red team runs](https://www.dataframer.ai/docs/api-reference/red-teaming/list-runs.md): Get all red team runs for your company
- [Get red team run](https://www.dataframer.ai/docs/api-reference/red-teaming/get-run.md): Get a specific red team run
- [Delete red team run](https://www.dataframer.ai/docs/api-reference/red-teaming/delete-run.md): Delete a red team run
- [Create an Anonymization Run](https://www.dataframer.ai/docs/api-reference/anonymization-runs/create.md): Start a new anonymization run to detect and mask sensitive data on your data
- [List Anonymization Runs](https://www.dataframer.ai/docs/api-reference/anonymization-runs/list.md): List all anonymization runs for your company, newest first
- [Get Anonymization Run](https://www.dataframer.ai/docs/api-reference/anonymization-runs/get.md): Retrieve an anonymization run by ID. Also use this endpoint to poll for completion.
- [Download All Anonymized Files](https://www.dataframer.ai/docs/api-reference/anonymization-runs/download-all.md): Get a presigned URL to download all anonymized files as a ZIP archive
- [Download Anonymized File](https://www.dataframer.ai/docs/api-reference/anonymization-runs/download-file.md): Get a presigned URL to download a single anonymized file from a completed run
- [Delete Anonymization Run](https://www.dataframer.ai/docs/api-reference/anonymization-runs/delete.md): Soft-delete an anonymization run so it no longer appears in list views
- [Rotate API key](https://www.dataframer.ai/docs/api-reference/api-keys/rotate.md): Revoke your current API key and receive a new one

## OpenAPI Specs

- [openapi.with-code-samples](/docs/openapi.with-code-samples.yml)

## Optional

- [Support](mailto:info@dataframer.ai)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.