Qatalog

Implements a unified search index across heterogeneous data sources (Salesforce, Tableau, Looker, databases, data warehouses) by crawling and cataloging metadata from each system's native APIs and connectors. Uses a centralized metadata repository with full-text search and semantic indexing to enable employees to find data assets without direct access to underlying systems or requiring data engineering expertise. The search interface abstracts away source-specific query languages and access patterns, presenting a single search box that returns results ranked by relevance and metadata enrichment.

Continuously polls or subscribes to metadata change events from connected data sources (databases, data warehouses, BI tools, SaaS platforms) and updates the central catalog in near-real-time. Uses source-specific connectors that translate each system's metadata schema (e.g., Salesforce custom fields, Tableau workbook structure, Looker explores) into a normalized internal representation. Implements change detection at the metadata level (schema changes, asset renames, ownership updates) rather than data-level changes, reducing computational overhead while keeping the catalog fresh.

Provides a shared interface where team members can add descriptions, tags, business glossary terms, and custom metadata to data assets without modifying source systems. Uses a lightweight permission model (owner, editor, viewer roles) to control who can modify asset metadata. Supports bulk tagging operations and template-based annotations to standardize metadata across similar assets. Changes are tracked with audit logs showing who modified what and when, enabling teams to maintain a living data dictionary that evolves with organizational knowledge.

Constructs a directed acyclic graph (DAG) of data dependencies by analyzing metadata relationships across sources (e.g., which Tableau dashboard uses which database tables, which ETL jobs feed which data warehouses). Supports both upstream lineage (showing source data) and downstream lineage (showing dependent assets). Provides interactive visualization of lineage chains and enables impact analysis queries (e.g., 'if this table is deleted, what breaks?'). Lineage is derived from metadata relationships and connector-specific dependency information rather than deep code/query parsing.

Provides a plugin architecture for building custom connectors to new data sources beyond the pre-built integrations (Salesforce, Tableau, Looker, etc.). Connectors implement a standard interface for metadata extraction (schema discovery, asset enumeration, ownership mapping) and are responsible for translating source-specific metadata formats into Qatalog's normalized schema. Includes SDKs and documentation for building connectors, with support for both pull-based (polling APIs) and push-based (webhooks) metadata delivery. Pre-built connectors for popular platforms are maintained by Qatalog; custom connectors are built and maintained by customers or partners.

Enables assignment of data stewards, owners, and subject matter experts to individual assets or asset collections, with role-based permissions controlling who can modify ownership and metadata. Supports bulk ownership assignment and automated ownership propagation (e.g., assigning a team as owner of all assets in a schema). Tracks ownership history and enables notifications to owners when their assets are accessed or modified. Integrates with identity systems (LDAP, SSO, directory services) to sync organizational structure and enable role-based access control.

Integrates with external data quality tools (e.g., Great Expectations, Soda, dbt tests) to display quality metrics and test results alongside asset metadata in the catalog. Pulls quality scores, test results, and anomaly detection alerts from quality platforms and displays them in asset detail pages. Enables filtering and searching by data quality status (e.g., 'show me all datasets with quality score < 80%'). Does not compute quality metrics itself; acts as a display layer for metrics generated by external tools.

Provides a free tier with limited features (basic search, single data source, limited users) that allows teams to test core cataloging functionality without upfront cost or sales process. Includes guided setup workflows that walk users through connecting their first data source, creating initial asset collections, and inviting team members. Uses a low-friction SaaS model with no installation or infrastructure setup required. Upgrade path to paid tiers is self-serve; customers can add data sources, users, and advanced features through the product UI without contacting sales.