Anomalo

Anomalo is an AI data quality and observability platform for enterprise data teams, run as SaaS or in the customer's VPC and sold by quote after a demo.

Anomalo watches the tables in a company's data warehouse and raises a flag when their contents change in ways nobody wrote a rule to catch. At its core sits a proprietary data profiling and prediction engine that learns each table's normal patterns and spots shifts across billions of rows. Anomalo, Inc. is independent and was co-founded by chief executive Elliot Shmukler and chief scientist Jeremy Stanley, both formerly of Instacart. The aim is to stop broken data before it reaches reports, dashboards and AI models. The work is split among agents running on that engine. The Table Observability Agent tracks freshness, volume, schema and distribution on every table without per-table setup. The Data Quality Rules Agent turns a check described in natural language into a rule, and a no-code interface offers a catalog of validation checks for anyone who prefers forms. The Proactive Data Insights Agent writes analyst-style reports on noteworthy changes without being asked, while the Conversational Analytics Agent, called AIDA, answers questions about the data in plain language. Column profiles show every column's expected values and how they have changed over time. Anomalo is built for enterprise data organisations. Data engineers rely on the machine learning monitoring so they do not maintain brittle rules table by table. Data stewards in business divisions can own their own monitoring through the no-code checks, without waiting on engineering. Companies that need their data quality tooling inside their own cloud account can run it fully in their VPC instead of as a hosted service. Every path to buying it starts with a demo request. A typical setting is a warehouse on Snowflake, Databricks or BigQuery, where late, shrunken or reshaped tables have to be caught before a morning dashboard goes out. Teams on Apache Airflow add its native operators to their DAGs, so a failed check halts low-quality data partway down the pipeline. Analysts who define business metrics in dbt hand those metrics to Anomalo to watch for unusual movement. Alerts arrive in Slack as issues are found. On the governance side, results show up in Alation's Table Overview and Health sections and as quality events in Atlan, so anyone who finds a table in the catalog can see whether to trust it.