Quick Start Guide

Welcome to Qualytics! This guide takes you from initial access to your first quality signals. You will connect data where it already lives, use Sync to map it, run Profile to understand its behavior and create most AI Managed checks, and run Scan to surface anomalies for review.

Deployment Access

Your Qualytics administrator provides the URL and sign-in method for your managed or self-hosted deployment. The interactive API documentation is available from the deployment at /api/docs.

Before You Begin

Confirm that you have:

• Access to your Qualytics deployment.
• Connection details and credentials for the first source datastore.
• Any network access required by the selected connector.
• A business process or critical dataset to prioritize.

Your platform role controls deployment-wide capabilities. Your team permission on a datastore controls what you can do with that data. See Onboarding for the full first-run workflow.

Signing In

The sign-in methods available to you depend on how your deployment is configured.

Method 1: Direct Credentials

Use the credentials provided by your Qualytics administrator when direct sign-in is enabled.

Method 2: Enterprise SSO

When enterprise SSO is configured, sign in through your organization's identity provider.

Getting Started Checklist

To produce your first quality signals:

1. Connect your first source datastore.
2. Confirm that Sync discovered the expected tables, files, and fields.
3. Run Profile with the appropriate AI Effort and Oversight settings.
4. Review the resulting AI Managed checks and add Authored checks for business expectations.
5. Run Scan and investigate the resulting quality scores and anomalies.
6. Configure Flows or integrations for the responses your team needs.

The sections below introduce each part of the workflow.

Understanding Datastores

In Qualytics, a source datastore represents data accessed through a configured connection. Supported connectors fall into two broad groups:

Database and Warehouse Datastores

Traditional relational databases (RDBMS), data warehouses, and analytical databases. These connect to Qualytics in one of two ways:

• JDBC datastores: connected through a standard JDBC driver (for example, PostgreSQL, Snowflake, or SQL Server).
• Native datastores: connected through the platform's own integration instead of a JDBC driver (currently available for Databricks).

File-Based Datastores

Files stored on cloud object storage, connected to Qualytics as DFS datastores (distributed file system):

• Cloud storage such as Amazon S3, Azure Blob Storage, and Google Cloud Storage.
• Raw files (CSV, XLSX, JSON, Avro, Parquet).

Connecting Your First Datastore

Adding a Source Datastore

1. On the welcome page, select Connect your data. From the datastore list, select Add Source Datastore.
2. Select a connector or reuse an existing connection.
3. Enter the connection and datastore settings.
4. Choose whether to Initiate Sync, then select Test connection.
5. Select Finish, or select Next to link an optional enrichment datastore.

Enrichment Datastores

A linked enrichment datastore can persist configured Scan outputs such as anomalies, source-record examples, and remediation data. Export and Materialize operations can write selected metadata or data snapshots. What is written depends on the settings and operations you use.

You can either:

1. Configure a new Enrichment Datastore.
2. Select an existing Enrichment Datastore from the dropdown.

Core Operations

After connecting your datastore, Sync, Profile, and Scan build the first set of governed quality signals.

1. Sync Operation

Sync establishes the inventory that later operations use:

• Detects new, changed, or removed containers and fields and updates the datastore inventory in Qualytics.
• Establishes the containers and fields available to Profile and Scan.
• Can start after creation when Initiate Sync is selected; otherwise, run it from the datastore.

2. Profile Operation

Profile records statistics and observed patterns for the selected data:

• Records profile metadata.
• Calculates measures such as:
  • Basic metrics (type, min/max, and lengths).
  • Advanced analytics (skewness, kurtosis, and correlations).
  • Value distributions and patterns.
• Generates and maintains most AI Managed checks according to the selected AI Effort level.
• Uses Oversight to determine whether new AI Managed checks become Active or remain Draft for review.

3. Scan Operation

Scan evaluates the active checks within the selected scope:

• Evaluates applicable AI Managed and Authored checks.
• Records failed checks as potential issues for review.
• When an enrichment datastore is linked, writes configured anomaly data and source-record examples there.
• Updates the related quality scores.

After a Scan completes, you can review:

• Start and finish times.
• Records processed.
• Anomalies detected.
• Quality scores.

Reviewing Data Quality

Quality Checks

Qualytics uses two types of quality checks:

1. AI Managed Checks

• Most are inferred from observed data during Profile operations. Volumetric and Freshness checks use recorded Observability measurements.
• Provide broad starting coverage without requiring every check to be written manually.
• Can become Active automatically or remain Draft for review, depending on Oversight.
• Qualytics maintains the check during later Profile operations while it remains Qualytics-owned.
• Marked with a purple AI badge (four-point star icon).

2. Authored Checks

• Manually created by users.
• Capture business rules and known expectations that observed patterns cannot provide.
• Range from guided rule types to advanced SQL expressions.
• Can be saved as check templates and applied to compatible data.

Platform Navigation

Explore Dashboard

Explore brings recorded quality signals and operation history together across the source datastores you can access:

1. Insights

• Recorded quality scores, active checks, and anomalies.
• Profile and Scan activity.
• Filter by Source Datastores and Tags, then select a Report Date and Timeframe.

2. Activity

• Review operation history, status, results, and logs.
• Use the calendar heatmap to see operation counts by day.
• Review and manage operation schedules.

3. Profiles

Browse profiled assets across datastores you can access:

• Containers, including tables, files, and computed assets.
• Active, masked, missing, and excluded fields.

4. Observability

Review volume, freshness, and field-level measurements:

• Volumetric and Freshness measurements recorded on their configured schedule.
• Metric measurements recorded by relevant Profile and Scan operations.
• Charts and anomalies produced when observability checks evaluate those measurements.

5. Checks

Unified view of all data quality validations across datastores:

• Use All, Active, Draft, or Archived; narrow Active and Draft checks by Important, Owned, Favorite, or All.
• Filter by Source Datastore, Tags, or Importance.
• View validation results by table and field, including pass/fail status and anomaly counts.

6. Anomalies

Review potential quality issues surfaced across datastores:

• Review open anomalies by Active, Acknowledged, Assigned, or All, and archived anomalies by resolution status.
• View the datastore, container, affected fields, failed checks, and detection date.
• Use severity, assignee, and status to prioritize investigation. Review aggregate trends in Insights.

Configure Responses and Access

Tags

Use shared business context to organize and prioritize work:

• Categorize assets with shared business context.
• Use tags in Explore filters, operation scope, and Flow conditions.
• Use weight modifiers to influence aggregate quality scores and anomaly priority.

Flows

Coordinate supported responses:

• Start from a schedule, manual run, supported operation, or matching anomaly event.
• Run supported operations, update anomalies, or send configured notifications, tickets, and workflow requests.
• Review execution status and timestamps.

Platform Settings

Access key configuration areas:

1. Connections

   • Manage reusable connection definitions.
   • Start adding datastores from a connection.

   

2. Integrations

   • Configure the AgentQ provider.
   • Configure supported data catalog, alerting, ticketing, and workflow integrations.

   

3. Defaults

   • Set deployment-wide defaults for Observability, Profile, and Scan.
   • Available to Admins.

4. Security

   • Manage users and Teams.
   • Manage platform roles and datastore team permissions.

   

5. Tokens

   • Manage Personal tokens and, where permitted, Service tokens for programmatic access.
   • Tokens can be revoked, restored, or deleted at any time.

   

6. Audit

   • Review and export platform activity.
   • Available to Admins.

7. Status

   • Review deployment configuration, component health, and capacity.
   • Available to Managers and Admins.

   

Next Steps

Now that you're familiar with the Qualytics basics, consider:

1. Setting up additional datastores.
2. Creating Authored checks for known business expectations.
3. Configuring Flows and notifications for supported events.
4. Reviewing quality scores and anomalies with the responsible data owners.

For detailed information on any topic, explore the relevant sections in our documentation.