Compass

LangGrant connects large language models to your production databases for trusted, governed data analysis. Instead of one-off answers, it produces persistent, versioned Data Plans that show their work - the steps and logic behind every result. Teams review, validate, and reuse these plans without moving data. A single question can join data across multiple databases, and LangGrant auto-adapts to each schema. It works with models like Claude, OpenAI, and Google Gemini, and supports Snowflake, SQL Server, Oracle, PostgreSQL, BigQuery, Redshift, Azure SQL, Databricks, and MySQL. Data Plans move through a code-like lifecycle (test, sign-off, staging, approved) with access controls, PII protection, compliance validation, and audit trails. By reusing reasoning instead of paying AI for every question, it cuts cost, improves collaboration, and builds organizational intelligence. From the team behind Windocks, recognized by Gartner.

nao is an AI-powered data IDE designed specifically for data teams, combining a code editor with native integration to your data warehouse so you can write, test, and maintain data-centric code with full context. It supports warehouses such as Postgres, Snowflake, BigQuery, Databricks, DuckDB, Motherduck, Athena, and Redshift. Once connected, nao replaces a traditional data-warehouse console by offering schema-aware SQL auto-completion, data previews, SQL worksheets, and the ability to switch easily between multiple warehouses. The core of nao is its AI agent, which has full awareness of your actual data schema, tables, columns, metadata, and your codebase or data-stack context. It can generate SQL queries or full data-transformation models (e.g., for dbt workflows), refactor code, add or update documentation, run data-quality checks and data-diff tests, and even surface insights or run exploratory analytics, all while respecting data structure and quality constraints.