GW
HomeWorkAboutNotesStackLab
Experience atOmnicom Group

Driving financial systems and automation across global teams.

Builder ofEnterprise AI Systems

AI agents, audit workflows, and decision intelligence at scale.

Based inNew York City

Building systems for enterprises around the world.

HomeNotesWorkAboutStackLabPartners
XLinkedInGitHubEmailllms.txt
© 2026 · New York City
All enterprise work

Decision intelligence case study

Real estate decision intelligence.

A governed real-estate portfolio app and Genie-based analysis surface for moving from scattered asset detail to a decision-ready portfolio view.

Databricks logoDatabricks AppsDatabricks logoDatabricks GenieReactNext.js
Ask the project copilotGrounded in this public-safe case study.

Business problem

What people needed to solve.

Real estate teams work from asset detail spread across leasing, valuation, and finance systems, so building a portfolio view means manual reconciliation and every conclusion is hard to trace back to a source. Leaders need one governed place to see which assets need attention and to ask follow-up questions without losing that source context.

Decision frame

Questions the work needed to answer.

  • Which properties or portfolio segments need attention?
  • What detail supports a portfolio-level conclusion?
  • How can users ask follow-up questions without losing the governed source context?
Interactive operating flowSelect any node for context or ask the AI guide.

Selected Operating layer

Governed portfolio app + Genie

Designed a Databricks-hosted React and Next.js application for portfolio, asset, and exception views, paired with a Databricks Genie path for natural-language questions over approved real estate data. Each answer keeps the underlying detail, the user context, and the source lineage visible rather than returning an untraceable figure.

Implementation

From the business problem to a working operating model.

Designed a Databricks-hosted React and Next.js application for portfolio, asset, and exception views, paired with a Databricks Genie path for natural-language questions over approved real estate data. Each answer keeps the underlying detail, the user context, and the source lineage visible rather than returning an untraceable figure.

  1. 01

    Organized approved real-estate detail into portfolio, asset, and exception views.

  2. 02

    Built a Databricks-hosted application for a master portfolio experience instead of a static reporting page.

  3. 03

    Added a governed Genie path for natural-language analysis and follow-up questions.

Controls, approvals, and delivery constraints

The work around the work.

Enterprise systems change only when data, access, testing, ownership, and evidence move together.

  • Defined which views were approved for analysis and how each audience could use them.
  • Kept portfolio detail, user context, and source lineage visible instead of presenting an untraceable AI answer.
  • Designed the experience around authorized access, not public data exposure.

Prior art

How leading teams approach this.

Public references for the same class of problem, so this work can be read against how other organizations are building it.

Morningstar

Morningstar Credit Analytics Launches AI Access to CRE Surveillance and CMBS Analytics

Morningstar exposed loan and deal level commercial real estate data to Claude over MCP so analysts query it in natural language while existing entitlement and access controls still apply, the same governed natural-language pattern used here.

Anthropic

Claude for Financial Services

Anthropic packages Claude for portfolio deep dives and diligence with connectors to governed data sources and full audit trails, and cites Bridgewater's AIA Labs using it for investment analysis, mirroring the explore-then-ask flow in this app.

Databricks

AI/BI Genie is now Generally Available

Databricks Genie answers natural-language questions grounded in Unity Catalog semantics and access policies, with named customers such as HP Inc and 7-Eleven, and is the Genie surface used in this case study.

Public-safe outcome

Illustrative of a portfolio intelligence workflow where a user explores the portfolio, inspects the supporting detail, then asks a governed question, with access scoped to what each audience is authorized to see.