dv01 launches AI Agents and MCP

Power your AI with dv01 Intelligence

Connect the AI tools and agents you already use to dv01’s governed data and analytics through one secure, permissioned connection. No new interface, no custom integration.

Model context protocol

Equip your AI with asset-backed finance expertise

General-purpose LLMs are powerful but aren’t built with the domain-specific context needed for structured finance workflows. dv01’s MCP bridges that gap.

  • Secure connection to dv01

    Bring dv01’s data and analytics into your AI applications and agents through a secure connection

  • Plug in and get set up in minutes

    Easily connect dv01 to your AI tools by logging in and authenticating your account. No custom integration or implementation required.

  • Semantic layer, built in

    Ensure your AI understands how metrics are defined and calculated, how they relate to one another, and how dv01’s data should be queried and analyzed — instead of approximating every time you ask a question.

Connect dv01 to internal AI Agents and major LLMs

  • Claude
  • ChatGPT

Make structured finance conversational

The dv01 MCP enables analysts to access, analyze, and model dv01 data using natural language across two core capabilities.

Collateral analysis

Find what’s driving collateral performance

dv01’s AI Assistants handle repeatable, document-intensive work inside the platform. Outputs remain editable and source-linked, so your team can review, refine, and complete the work.

Use the dv01 MCP to:

  • Query the deals, portfolios, and loan-level data you have access to
  • Stratify data by loan, borrower, and asset-class attributes
  • Analyze performance trends across time, vintage and loan age
  • Identify the individual loans behind any result
Agentic Collateral Analysis
Cashflow modelling

Model and stress-test cashflows under different scenarios

Use the dv01 MCP to connect supported AI applications and internal agents to licensed dv01 data and analytical tools. Ask questions, run analyses, and receive structured results without building a custom integration or adopting another interface.

Use MCP for:

  • Project cashflows for deals and loan pools
  • Model scenarios by performance, pricing, and rate assumptions
  • Analyze waterfall results across collateral, bonds, and tranches
AI chat powered by the dv01 MCP comparing a Class A note's base case with a 1.5 CDR / 8 CPR stress: weighted average life, yield, z-spread and modified duration

One trusted foundation for AI-driven analysis

AI is only as reliable as the data and tools it has access to. The dv01 MCP ensures you get answers you can rely on, every time.

Three stacked layers. dv01 MCP: natural language in, structured analysis out, with provenance intact and no dashboard in between; available for Claude and ChatGPT. Semantic layer: aggregates raw loan-level fields such as balances and payment history into the metrics that matter, like CPR, CDR and cumulative loss, with one governed definition every time. Data layer: 120M+ loans, 900+ securitizations and $5T in original balance, normalized into one warehouse and one schema across every asset class.
  • Ground every answer in dv01 data

    Your AI works from dv01’s standardized, AI-ready loan-level data, so every output starts from the same foundation.

  • Give AI structured finance context

    dv01’s semantic layer helps AI identify the relevant data, relationships, and tools for each query, without relying on the model to interpret raw data on its own.

  • Apply the same logic across queries

    Centrally defined calculations and methodologies apply the same logic to the same inputs to keep results consistent.

  • Understand and verify every result

    Permissioned access, source traceability, editable outputs, and human review where it matters.

Ready to put dv01 AI to work?

Contact us