Overview
Secure, live access to your nCino environments for AI assistants, using the Model Context Protocol.
Early access. The nCino MCP is provided "as is" while in early access. Features, availability, and timelines may change at nCino's discretion and should not be relied on for purchasing decisions.
The nCino MCP connects an AI assistant, such as Microsoft Copilot, Claude, ChatGPT, Gemini, or Perplexity, directly to your nCino environments. Once connected, you can ask questions about your configuration in plain language and get answers from the live system instead of digging through setup screens. You can also ask it to make changes in a sandbox, like turning on a feature flag or adding a picklist value.
It uses the Model Context Protocol, an open standard for connecting AI assistants to external systems. It is available to nCino customers and their implementation partners.
What you can do with it
- Understand your configuration. Ask what validation rules, flows, approval processes, Apex, page layouts, permission sets, or picklist values exist in an org, and why something behaves the way it does.
- Explore the data model. Look up fields on an nCino object, see how objects relate to each other, and compare how an object changed between releases.
- Plan releases and upgrades. Find nCino releases and package versions, see which release an org is on, build the upgrade path to a target release, and work out install order and dependencies.
- Make changes in a sandbox. Turn feature flags on or off, update system properties, deploy declarative metadata such as picklist values, page layouts, and permission sets, and create or update records.
- Test Apex. Run Apex tests in a sandbox and check code coverage.
- Search nCino documentation. Search help articles, known issues, and technical documentation, and get the full article back.
What to read next
- Configure: what your institution needs in place, and how to add the connector in your AI assistant.
- Tools: everything the connector can do.
- How to use: prompts to try.
- Limitations: what it will not do, and what to expect during early access.
- Support: getting help, giving feedback, billing, and privacy.
Updated about 1 month ago
