AI CRE tips: how investment managers use AI across the deal lifecycle

A GP is three weeks from a fund close when an LP asks a straightforward question: What’s the current occupancy across the portfolio, broken out by asset type? The answer exists, but it’s scattered across four spreadsheets, two of which haven’t been updated since the last distribution.

It’s a small request that exposes a much bigger problem. In commercial real estate, critical information is often fragmented across spreadsheets, systems, and teams, turning routine questions into time-consuming exercises in finding, checking, and reconciling data.

And that friction doesn’t stop at investor reporting. It follows CRE professionals through nearly every part of the job, from sourcing and underwriting deals to raising capital, communicating with investors, and managing the finance operations behind it all. This piece looks at practical ways to reduce that burden across each of these workflows, especially for teams already stretched across too many spreadsheets. 

What does AI in CRE actually mean for investment managers?

For an investment manager, artificial intelligence in commercial real estate isn’t a chatbot bolted onto a website. It’s software that reads a rent roll, flags the deals worth a closer look, drafts an investor update, or reconciles a distribution waterfall, using the firm’s own data instead of a template. Commercial real estate AI, in this sense, is less a single tool than a layer that sits across the workflows a firm already runs. The shift is from software that stores information to software that acts on it.

AI CRE workflow vs. traditional CRE workflow

The difference shows up clearest side by side.

Task	Traditional workflow	AI-enabled workflow
Deal screening	Analyst manually reviews every incoming deal against firm criteria	AI systems pre-screen deal flow and rank fit before an analyst opens the file
Underwriting	Analyst builds a DCF model from scratch over one to two days	AI generates a first-pass model in minutes, analyst refines assumptions
Investor onboarding	Subscription documents, accreditation, and KYC/AML handled as separate manual steps	One connected workflow moves an investor from interest to funded in days, not weeks
Reporting	Finance team builds LP updates by hand each quarter	Updates generate automatically from live portfolio and fund data

What is pushing CRE investment managers toward AI right now

Adoption isn’t happening because AI is fashionable. Four pressures are converging at once.

Most commercial real estate firms are already exploring or piloting AI. JLL’s 2025 Global Real Estate Technology Survey found that AI pilots among commercial real estate occupiers jumped from below 5% to 92% in three years, and investment firms are following the same curve. Firms past the pilot stage are evaluating more deals with the same headcount, since screening and underwriting no longer eat an analyst’s whole week.

Manual back-office work is the cost a firm can cut fastest. Real-time portfolio data and same-week reporting are becoming the baseline expectation, not a differentiator.

Screen deal flow against your own criteria before an analyst opens the file. Feed incoming deals through an AI system that scores them against the firm’s return thresholds, asset type, and geography, so analysts only spend time on deals that already clear the bar.

1. Automate the first draft of a DCF model. What used to take an analyst a day or two of assumption-building can be generated in minutes, leaving the team to sanity-check inputs and stress-test scenarios instead of building the model from a blank sheet.
2. Apply AI-powered market trend analysis to catch shifts early. Pattern recognition across pricing, absorption, and lease data can flag an emerging submarket before it shows up in a quarterly report.
3. Layer predictive analytics onto data you already have. The same rent rolls and comps a firm has been collecting for years become a far better forecasting input once a model can find the patterns a human would miss on a spreadsheet.

AI tips for fundraising and LP onboarding

Fundraising is where slow, manual processes cost a firm the most in lost time with prospective investors, and where the right AI tools show up fastest in a GP’s day-to-day.

1. Run subscriptions, accreditation checks, and KYC/AML from one workflow. Instead of three separate systems and three separate document trails, a connected process moves an investor through every step without re-entering the same information twice.
2. Cut investor onboarding from days to minutes. Automated accreditation and document verification remove the back-and-forth that used to stretch a simple subscription into a week of email chasing.
3. Replace a spreadsheet-based pipeline with a live fundraising dashboard. A GP raising from 200 LPs across three vehicles needs to see where every commitment stands without opening a new tab for each fund.
4. Personalize outreach at scale. AI-powered CRM and email tools let a small IR team send LP-specific updates to hundreds of investors without writing each one from scratch.

AI tips for investor relations and reporting

LPs increasingly expect the kind of instant access they get from their bank.

AI capability	What it replaces	What LPs get
Automated, LP-specific reporting	A finance team drafting each update by hand	Updates that reflect each investor’s specific commitments and returns
Automated waterfall processing	Manual distribution calculations in a spreadsheet	Faster, more accurate distributions with a clear audit trail
24/7 investor portal access	Static reports sent on a quarterly schedule	Performance data available whenever an LP wants to check it
AI-generated K-1s and tax documents	A finance team producing tax paperwork by hand at scale	Documents delivered faster, without added headcount

AI tips for investment management and finance operations

Behind the LP-facing side of the business, finance and asset management teams are dealing with their own version of the same problem: too much manual reconciliation, spread across too many systems. The AI technology behind these workflows is the same pattern-matching that shows up everywhere else in commercial real estate.

1. Run real-time dashboards for NOI, occupancy, and market trends. Instead of pulling numbers together for a monthly review, the numbers are already there, updated as the underlying data changes.
2. Use predictive analytics on refinancing and exit timing. Modeling rate scenarios and market conditions ahead of time gives a firm more room to act instead of reacting to a maturity deadline.
3. Automate capital calls and distribution workflows. A finance team that used to spend two days on a capital call notice can generate, send, and track it in an afternoon.
4. Centralize fund data across every deal. Spreadsheet-based portfolio tracking breaks down once a firm passes a certain number of assets; a single source of data doesn’t.

AI prompting tips for CRE professionals

Getting good output from an AI system depends less on the tool and more on how the CRE professionals using it ask for what they need. The same discipline applies whether someone is chatting with a generative AI assistant or setting up an AI agent to run a task on a schedule.

1. Give the AI relevant property and market context. A prompt that includes the asset type, submarket, and specific numbers will produce a far more useful answer than a generic question about “commercial real estate trends.”
2. Be specific about the output you want. “Summarize this rent roll” gets a summary. “Flag every unit rolling in the next 12 months and estimate the rent gap” gets something you can act on.
3. Use AI to analyze, not just generate. The tools are as good at spotting a pattern in a data set as they are at drafting a paragraph, and that analytical use tends to get skipped.
4. Refine your prompts with follow-up questions. Treat the first answer as a starting point, not the final one. Asking “what would change this conclusion” often surfaces the assumption that mattered most.
5. Build reusable prompts for tasks you do every week. A prompt template for underwriting summaries or investor update drafts saves the rebuilding work every single time a similar request comes up.

Where AI adoption goes wrong in CRE firms, and why it keeps happening

Not every commercial real estate AI rollout delivers on its promise, and the failures tend to follow a pattern. 

1. Adopting point solutions that don’t connect to the rest of the workflow: One GP replaced its manual underwriting spreadsheet with a standalone AI tool, then spent more time exporting outputs into its CRM and reporting systems than it had saved on the underwriting itself.
2. Manually recreating data that already exists somewhere else: When deal data lives in one system, investor data in another, and fund accounting in a third, a firm ends up re-entering the same numbers instead of connecting the systems that already have them.
3. Underestimating how fast investor expectations are moving:. A firm that’s comfortable with quarterly PDFs is going to have a hard conversation with an LP who’s used to checking a portfolio balance from their phone.
4. Treating AI as a one-time project instead of an ongoing shift: A model trained on last year’s data and never revisited stops reflecting how the firm actually operates within a few quarters.

How Cortex brings AI into every Agora workflow without adding a separate tool

Most firms don’t need another standalone AI product bolted onto their stack, on top of the CRM, the portal, and the fund accounting system they already run. Cortex is Agora’s AI work surface, built directly into the platform a firm already uses to manage deals, investors, and fund data. 

Rather than asking a team to copy information into a separate AI system, Cortex works from the data already sitting in Agora: a firm’s own deal history, investor records, and fund performance. Ask it to draft an investor update and it pulls from that fund’s actual numbers. Ask it to flag deals that fit a firm’s criteria and it’s screening against the firm’s own past decisions, not a generic model. The result is one place to work instead of another tab to manage. For firms that want to go further, Agora’s API also lets a firm connect its own AI agent to Agora’s entity and transaction data, instead of waiting on a manual export.

How Agora applies AI across the full CRE investment management lifecycle

Agora supports more than 150,000 investors and $300 billion in assets under management for firms that raise external capital. That scale means the platform touches every stage where the AI tips in this article apply, and it’s built so a firm doesn’t need to stitch together a separate AI system for each one:

• Cortex, Agora’s AI work surface — drafts investor updates, screens deals, and answers questions using a firm’s own historical data rather than a generic model.
• Deal sourcing and underwriting — AI-assisted screening and first-pass modeling run against a firm’s own return criteria, inside the same CRM used to track the deal afterward.
• Fundraising and investor onboarding — subscriptions, accreditation, and KYC/AML move through one connected workflow instead of three disconnected systems.
• Investor relations and reporting — an investor portal with real-time performance data and automated, LP-specific report generation, so updates don’t wait on a manual quarterly build.
• Fund accounting and distributions — automated waterfall calculations, capital call notices, and K-1 generation, produced from the same fund data used everywhere else in the platform.
• Bookkeeping center — automated transaction categorization and reconciliation across entities, keeping the books current without a separate manual close process each month.

The AI tools for real estate investors built into Agora’s CRM, investor portal, and fund accounting handle the deal lifecycle end to end, from the first screen on a new deal to the K-1 an investor receives three years later.

The next move for CRE firms weighing AI

None of this requires ripping out what a firm already uses. The firms getting real value from AI in commercial real estate started with one workflow, and expanded from there once it worked. The commercial real estate firms still waiting are the ones most likely to be reacting to an LP’s question instead of already having the answer.

The bigger risk is picking a point solution that solves one problem while creating three more integration headaches. Firms that connect AI to data they already have, instead of building a parallel system, are the ones seeing it stick.

If you’re weighing where to start, talk to an expert about what a firm your size and structure has seen work first.