Skip to content

Published November 2026 · Independent research; not peer reviewed

Practical companion · Issue 01

AI Opportunity Explorer

Start with the problem, not the tool. 28 task-first use cases, each with the workflow change, the people who stay in the loop, and what to measure.

Companion to Issue 01, What AI Actually Does at Work and Issue 02, The AI Implementation Gap. Real estate appears as a recurring lens, alongside other industries the series covers.

28 of 28 use cases

  • Documents & paperwork

    Invoice and receipt capture

    Illustrative designWidely available
    Operational problem
    Vendor invoices are keyed in by hand and coded inconsistently.
    Type of AI solution
    Document extraction plus suggested coding: read invoice fields and propose account and job codes.

    Before

    Clerk types vendor, amount, date and code from each invoice.

    After

    Fields and codes are pre-filled; clerk approves or corrects; exceptions queue for review.

    Human checkpoints

    • Approval thresholds unchanged
    • New vendors always reviewed
    • Duplicate detection reviewed weekly

    Data you need

    • Historical coded invoices
    • Chart of accounts / job codes
    • Vendor master
    Deployment pathway
    Configure
    Set up templates, rules, prompts or permissions inside a tool you already use.
    Complexity
    Low

    Safeguards

    • Payment controls stay with people
    • Watch for vendor-impersonation fraud

    KPIs to baseline, then measure

    • Invoices processed per hour
    • Coding correction rate
    • Days to approve

    A small pilot

    1. One entity or project for one month
    2. Track corrections by field

    Real estate lens

    Property-level payables and construction subcontractor invoices.

    Evidence status: Illustrative design

    Design pattern for planning. No published outcome is cited; measure your own baseline.

Selected: Invoice and receipt capture

About this directory

  • No vendors or products are named or endorsed. Each card describes a type of solution.
  • No performance, savings or ROI figures are claimed. KPIs are measures to baseline yourself before and after a pilot.
  • "Survey-reported task use" means the Census BTOS lists the task category among reported generative AI uses. It says nothing about quality or results.
  • Published results apply to the settings studied and do not predict outcomes elsewhere.
  • Industry filters reflect sectors the series already covers; "any industry" cases appear under every filter.
  • Independent research; not peer reviewed. Not legal, compliance or investment advice.

Deployment pathways

Buy:
Use an off-the-shelf tool largely as delivered.
Configure:
Set up templates, rules, prompts or permissions inside a tool you already use.
Integrate:
Connect an AI service to your systems of record (CRM, accounting, property management).
Custom build:
Design and maintain your own workflow, models or data pipeline.

Sources

  1. Brynjolfsson, E., Li, D., & Raymond, L. R. (2025). Generative AI at Work. Quarterly Journal of Economics, 140(2), 889–942. https://doi.org/10.1093/qje/qjae044
  2. Dillon, E. W., Jaffe, S., Immorlica, N., & Stanton, C. T. (2025). Shifting Work Patterns with Generative AI. NBER Working Paper 33795. https://www.nber.org/papers/w33795
  3. U.S. Census Bureau. BTOS Core and AI Content (official questionnaire). https://www.census.gov/hfp/btos/downloads/BTOS%20Core%20and%20AI%20Content.pdf
  4. U.S. Census Bureau. Business Trends and Outlook Survey, 2026 AI supplement tables. https://www.census.gov/hfp/btos/data_downloads?so=1
  5. Holmes, N., Sanchez-Moyano, R., & Simms, S. (2026, March 23). Early Findings on Small Business Use of AI. Federal Reserve Small Business Credit Survey analysis. https://www.fedsmallbusiness.org/analysis/2026/early-findings-on-small-business-use-of-ai