Connect Addepar to your AI system with MCP (closed beta)
The Addepar MCP connects your portfolio data and intelligence with your AI systems, so you can ask questions about your entire book of business wherever you work. The MCP also includes Addepar's specialized ownership structure, calculation logic, and permission-aware access, so any system it’s connected to gets the full context of how your portfolios are structured and calculated.
Analyze a household’s holdings, market value, asset allocation, and performance. Understand alternatives exposure across capital calls, distributions, NAVs, and unfunded commitments. Review transactions, identify trends across clients, or gather the data needed to prepare for a quarterly review.
Note: The Addepar MCP is currently in closed beta. Contact your Addepar representative to request access.
In this article:
- What is the Addepar MCP?
- How MCP compares to APIs
- How the Addepar MCP compares to Addison
- Connect the Addepar MCP to your AI system
- Combine Addepar data with other sources
- Example use cases
What is the Addepar MCP?
Model Context Protocol (MCP) is an open standard that lets AI systems discover and use approved capabilities from other platforms. The Addepar MCP is a connector that lets you access your portfolio data from any AI system without switching windows or writing code.
Every answer the Addepar MCP provides is backed by the same calculations engine your team relies on. Performance, attribution, exposure, and transaction data all come through with the same precision and methodology you'd see in Addepar directly.
Your Addepar permissions and portfolio access carry through the connection. The Addepar MCP can only access data you're already allowed to access in Addepar.
How MCP compares to APIs
With APIs, a developer writes code to connect two systems. With an MCP server, you authorize a connection between an AI system and another platform. The system reads what capabilities are available and uses them directly. You ask a question in plain language, and the system pulls the right data to answer it.
| APIs | Addepar MCP | |
| How to start | An engineer reads documentation, then builds an integration | Your AI system connects to available capabilities automatically |
| Typical time to set up | Weeks or months | Connect in minutes, then ask questions immediately |
| What it returns | Raw data that your custom app displays | Data with the context of Addepar’s calculations and ownership structure, delivered in plain language |
| User permission and portfolio access | You manage in your custom app | Addepar's permissions and portfolio access are carried through automatically |
How the Addepar MCP compares to Addison
Addison and the Addepar MCP have the same foundational intelligence, but serve different purposes. You might use Addison when you're working in Addepar, and use the Addepar MCP when you want to bring Addepar intelligence into your broader workflow. Both are permission-aware, so you can only access data that you’re allowed to access in Addepar.
| Addison | Addepar MCP | |
| Who it’s for | Anyone who wants quick answers about their book of business | Anyone who wants to connect Addepar data with data from other platforms |
| Setup and tech required | No setup––already built into Addepar | Connect in minutes |
| Where it’s used | Addepar | Your AI system |
| What it can access | Addepar data and Addepar Learn | Addepar data, Addepar Learn, and any other platform you connect via MCP |
Connect the Addepar MCP to your AI system
The Addepar MCP works with MCP-compatible AI systems. This currently includes Claude Desktop and Claude Code, with more coming soon.
To learn how to connect the Addepar MCP and participate in the closed beta, reach out to your Addepar representative.
Combine Addepar data with other sources
Get a complete view across your data ecosystem when you connect additional MCPs. With your portfolio data connected to the rest of your tech stack, you can get answers about the full picture in one place.
For example, if your other tools offer MCP:
- Pull from your benchmark provider to combine Addepar returns with benchmark data in a single attribution narrative.
- Pull from your research or note-taking platform to pair Addepar commitment data with meeting notes for a complete investment decision audit.
- Pull from your custodian to reconcile Addepar alternatives valuations against custodian snapshots automatically.
- Pull from your calendar to prepare portfolio summaries for each upcoming client meeting automatically.
- Pull from your CRM to combine Addepar portfolio data with client relationship history for a complete picture before any conversation.
Example use cases
Portfolio monitoring and performance
Ask your AI system to summarize what happened across your portfolios over a given period. It can identify material changes, explain performance drivers down to the position level, and flag accounts that need attention.
- "Give me a quick summary of the Thompson portfolio, including recent changes and anything I should flag before our meeting on Friday."
- "Which of my clients had the largest drawdown last month?"
- "What's the total AUM across all accounts I manage?"
- "What drove performance in the Smith household last quarter, down to the position level?"
Exposure tracking
Check how your clients are allocated across asset classes, sectors, or geographies. Your AI system uses Addepar's lookthrough capabilities to show exposure through fund structures and alternatives.
- "What is my total exposure to technology across all client accounts?"
- "Show me which clients have more than 30% allocated to alternatives."
- "How has the Garcia family's sector allocation shifted over the past year?"
Capital call and liquidity management
Ask about upcoming capital calls, identify available cash across accounts, and determine whether your clients have enough liquidity to cover upcoming commitments for your private fund investments.
- "What capital calls are coming due in the next 30 days across my book?"
- "Do my clients have enough cash to cover their upcoming commitments?"
- "Show me all available sources of cash across the Patel household accounts."
Data operations
Ask about data quality issues, recent imports, or account connections. Your AI system can help you spot problems and understand what needs attention without navigating through multiple screens.
- “Are there any accounts with missing returns for the last month due to incomplete data?”
- “List all accounts where the cash position is negative or showing a data error.”
- “Which households have uncategorized assets that need a ‘Model Type’ assigned?”
- “List all alternative asset positions that haven't been updated in the last 90 days.”
- “List all positions where the price is older than five days.”