Introduction to MarcoPolo

MarcoPolo is a secure connectivity layer that acts as a bridge between your AI Assistant and your diverse data ecosystem. It allows you to interact with, query, and analyze data from databases, cloud data warehouses, object storage, and SaaS applications using natural language, directly from within your favorite AI chat interface.

Instead of wrestling with different query languages or manually exporting CSVs, you can simply ask your AI assistant questions. The Large Language Model (LLM) powering your assistant uses its vast knowledge of SQL dialects, SaaS APIs, and object models to formulate the correct query. MarcoPolo then provides the secure, sandboxed environment where the LLM can execute that query against your data. The results are returned to your workspace, often cached in a high-performance in-memory database (DuckDB), allowing for rapid, iterative analysis and empowering you to make data-driven decisions at the speed of thought.

Who is it for?

MarcoPolo is designed for technically savvy knowledge workers who are comfortable using modern AI assistants and want to leverage them for data-related tasks. Our users typically fall into these groups:

• Business Users & Knowledge Workers: Product managers, operations leads, and marketing analysts who need to self-serve insights from company data without waiting for a data team.
• Data Teams: Analysts, Analytical Engineers, and Data Scientists who want to accelerate their workflow, automate repetitive data tasks, and explore data more fluidly using AI agents.

How it Works: The MarcoPolo Workspace

At the heart of MarcoPolo is the concept of a secure, dedicated Workspace. When you connect to MarcoPolo for the first time, we automatically provision a private "computer" for you in the cloud (specifically, a dedicated Kubernetes pod).

This Workspace is completely isolated and acts as your personal sandbox for all data interactions. Here is what makes it powerful:

• Secure by Design: Your data source credentials and configurations are securely stored and only mounted into your personal Workspace when it's active. This ensures your sensitive information is never exposed to the AI model or other users.
• Stateful Sessions: Your Workspace persists between queries. Any files you create, analyses you run, or results you cache remain available to you throughout your session, enabling complex, multi-step data workflows.
• Instant Updates: When you (or a team member) add a new data source or share an existing one, the changes are propagated to your Workspace instantly, ensuring you always have access to the latest resources without needing to restart or reconnect.

This architecture provides the security and isolation of a traditional data environment with the speed and flexibility required for modern AI-driven analysis.

For a more detailed look at the underlying mechanics, see the Technical Deep Dive section.