An investment firm gets answers without chasing every department.

The client is a full-service REIT with $50B in managed real estate assets, spanning equity investment across major property types, lending from senior to subordinate debt, development and asset management. The team needed to connect deal history, approved commitments and current asset performance across the business while protecting sensitive investment and financial records.

One investment question can reach across the business.

A partner asks why an investment has missed its approved plan. Acquisitions holds the original underwriting, development knows what changed, asset management has current performance, and the credit team tracks loan terms, covenants and downside exposure.

The evidence sits across shared drives, CRM records, spreadsheets and email. Properties appear under different names, and current models sit beside outdated copies. Finding a document is only the first step; the team still needs its date, approval status and relationship to the decision.

Across a platform of this size, the same history may be reconstructed for investment reviews, loan monitoring, development approvals and portfolio updates. Senior staff spend time supplying context that is already recorded elsewhere.

Mason connected that context and kept it current, so each review started with relevant evidence and unresolved questions.

Start with decisions and the evidence behind them.

Mason started with recurring questions: what changed since approval, which similar deals performed well, and had the firm considered an opportunity before? For credit teams, the review also connected borrower information, debt seniority and approved loan terms.

Each question was traced to its source records, decision owners and review steps. The work identified outdated models, inconsistent property names and decisions buried in email, then defined the systems, teams and records covered by the first release.

Connect the history before the next review.

The company brain links properties, deals, borrowers, companies and decisions across file systems, CRM and other business sources. It associates records with the relevant entity, version and date while source files remain in their existing systems.

Original underwriting stays separate from the current forecast. A review package brings together the approved assumptions, subsequent decisions and latest results, with source references and conflicting information visible to the reviewer.

Access followed source permissions and the user’s role across investment, credit and operating teams. Sensitive records stayed subject to the permissions governing the underlying sources.

Reviewers spent less time reconstructing documented history and more time investigating material changes. Missing decisions and unsupported explanations remained questions for the relevant people.

Turn connected knowledge into defined workflows.

The company brain supported deployed workflows with their own sources, business rules and human approvals:

• Investment and credit reviews: assemble prior underwriting, borrower records, loan terms and current asset performance into a source-linked package.

• Development approvals: compare proposed scope, quantities and rates with existing commitments before a person approves new work; separate true duplicates from legitimate changes.

• External diligence: reuse current property history and prior evidence to narrow repeated data-gathering assignments while preserving required technical, legal and reliance work.

• Follow-on AI workflows: reuse entity records, source connections and permissions across portfolio reporting, investor updates and other approved projects.

Reviewers used the connected evidence to remove duplicate commitments and correct errors before approval. Reuse also reduced repeated adviser work and integration spending. Financial benefits reflect changes to commitments and spending; people retained responsibility for approvals.

The workflow also reduced AI-processing costs. Against $2M in eligible annual production inference spending, a 35% reduction saved $700K. After $100K in additional retrieval, indexing and refresh costs, net savings were $600K.

ROI

Estimated annual cost avoidance is $3.65M: $3.05M in recurring benefits and $600K from six follow-on AI projects. Staff capacity is excluded, and shared implementation and operating costs still need to be deducted.

Approval benefits cover the firm’s share of $1.5B in annual commitments and exclude issues existing controls would catch. Integration savings apply only in the year each project is delivered.

40% less review-package preparation time

The team reduced review-package preparation time by 40%. Connected underwriting, borrower records and current asset information reduced the work of assembling a review-ready package. This released staff capacity is separate from the $3.65M in cost avoidance.

25% shorter follow-on delivery time

Follow-on workflows reached production 25% sooner. Reusing source connections, permissions and business context shortened setup while security checks, testing and adoption remained part of delivery. This reduction in delivery time is separate from the integration cost savings in the table.