Bruin

Bruin is an AI data analyst agent built for data teams that need verified answers to business questions inside Slack, Microsoft Teams, Google Chat, WhatsApp, Discord, Telegram, email, and the browser. The differentiator versus a general AI assistant is data integrity enforcement: Bruin reads live pipeline state, schema metadata, and data quality status before responding, and it declines to answer when the underlying data is stale or broken. Every team member receives the same answer from the same source. Bruin is built by Bruin Data Limited, a company incorporated in England and Wales, and serves data engineering and analytics teams at mid-market and enterprise B2B technology and SaaS companies. The open-source Bruin CLI handles ingestion, SQL and Python transformation, quality checks, lineage, and orchestration, and the managed Bruin Cloud layer adds the AI analyst, dashboards, and governance on top.

Bruin's integrations directory lists 203 built-in connectors, with 159 sources and 39 destinations shipped in its open-source ingestr tool, and Bruin states it builds a missing connector within 7 days from shared test credentials. Destinations include Snowflake, Google BigQuery, Databricks, Amazon Redshift, PostgreSQL, MySQL, Microsoft SQL Server, ClickHouse, DuckDB, Trino, Elasticsearch, IBM DB2, and Amazon S3. Oracle Database, Tableau, Looker, and Power BI are published as sources, so Bruin reads from those systems rather than writing dashboards back to them. Bruin publishes its own MCP server on a dedicated product page, which lets Cursor, Claude Code, and other MCP-compatible editors query databases, compare tables, ingest data, and run pipelines through the Bruin CLI. A Cloud MCP endpoint is documented separately for the managed platform.

Bruin publishes three plans. Open Source is $0 forever and self-hosted, covering the transformation engine, connectors, data quality checks, lineage and Dashboards as Code. Team is pay-as-you-go at $0.01 per credit with no seats and no minimum, and starts with $100 in credits plus 50 AI tasks without a credit card. Compute is billed per second while an asset runs, from 250 credits an hour ($2.50) on the default 256 MB size up to 16,000 credits an hour ($160) on 16 GB, and an AI task costs 100 to 300 credits ($1 to $3). Usage beyond the free credits is billed pay-as-you-go with no hard cap, and purchased credits do not expire. Scale is priced on request, with committed credits and volume discounts.

What Team does not include sits on Scale: SSO and SAML, bring-your-own LLM credentials and model restrictions, BYOC with data residency and private networking, dedicated egress IPs, the Cost Explorer, dedicated support with an SLA, and custom contracts with annual commitment. Team keeps AI conversations and SQL queries for 30 days, while Scale sets custom retention.

Bruin suits data teams that want ingestion, transformation, quality checks, and an AI analyst on a single warehouse-native platform. Teams weighing the same ground also look at Otto, which works on Apache Airflow orchestration rather than on the pipeline itself, and at Hex and Sigma, which pick up once the data has landed in the warehouse. Teams whose need is app-to-app automation rather than warehouse-native pipelines should look at n8n or Workato. Teams that want to drive data work from an AI editor rather than from chat can reach the Bruin CLI through its MCP server from Cursor and Claude Code. Teams that want an AI data analyst from Alibaba Cloud working across more than 50 data engines should consider Quick BI.

Current state Q4 2026: Bruin now publishes a pricing page, with a pay-as-you-go Team plan at $0.01 per credit alongside the free open-source engine and a custom Scale plan. September releases added reverse ETL destinations, CDC to BigQuery and Snowflake, partitioned backfills and read-only connections in the CLI, and dashboard version history, folders and pivot tables in Cloud. Bruin Data Limited holds ISO/IEC 27001:2022 certification issued 19 August 2026 by Prescient Security LLC, certificate 123371, valid to 18 August 2029, and is examined annually for SOC 2 Type 2 by Prescient Assurance. The bruin-data/bruin CLI repository holds 1.8k GitHub stars under the Apache 2 license, and the vendor homepage states 6.5k stars across all Bruin Data repositories. The vendor states that customer data is never used to train any model and that LLM calls run through enterprise endpoints under no-training agreements. Infrastructure runs on Google Cloud Platform and DigitalOcean.