Skip to main content

Prismatic & AI

Prismatic embraces AI in a variety of ways - from helping you and your customers build better integrations to enriching data within your flows and giving additional context to chat bots in your app.

Prismatic's adoption of AI focuses on four main areas:

  1. Code-native build tools to help AI coding agents build better integrations faster.
  2. Embedded workflow builder copilot to assist your customers as they build workflows in your app.
  3. Agentic flows as MCP tools to expose your integrations as MCP tools for an AI agent in your app.
  4. AI-enhanced data in flows to enrich data streaming through your flows.

Code-native build tools​

AI coding agents like Claude, GitHub Copilot or Codex are enhanced by Prismatic's MCP dev server and Prismatic Skills.

The MCP dev server adds tools to AI coding agents that allow them to connect to Prismatic's API and do things like "add the Salesforce connector to my integration" or "publish my integration and run a test of this flow".

MCP dev server tools

Prismatic Skills are Claude skills (agent skills) give your AI assistant context and examples of what "good" looks like for custom connectors and code-native integrations. So, when you ask your AI assistant to "write a trigger that polls new lead records from Acme", or "build a JSON Form field mapper for Salesforce", your agent has context for how to do that in a way that works well with Prismatic's platform and follows best practices.

Prismatic Skills example

Embedded workflow builder copilot​

The embedded workflow builder includes a copilot that gives your customers AI assistance as they build workflows in your app. The copilot can help your customers build workflows faster using natural language prompts.

Workflow builder copilot

Agentic flows as MCP tools​

You can mark some of your integrations' flows as "agentic" - meaning they can be called by an AI agent to perform a task. For example, you could have an "Update CRM" flow that your customers can call from a chat bot in your app to update a record in their CRM - all within the context of a conversation in your app.

This enhances the capabilities of your app's chat bots - they'll have access to integrations that your customers have enabled and can perform tasks that interact with the other apps and services they use.

Agentic flow example

AI-enhanced data in flows​

You can reach out to popular LLM providers like OpenAI or Anthropic from your code-native or low-code flows. Your flows can gather information from multiple sources using our library of built-in connectors, then call out to an LLM to summarize, analyze or enrich that data before taking an action.

Several examples are available - see Data Enrichment with AI.

AI enrichment flow