Introduction
We are in the midst of a commercial real estate technological revolution. AI is changing things fast. No longer is it just about writing emails. Love it or hate it, the firms who win in the fast paced world of CRE have adopted AI with open arms. In this article by the Cindtoro team we are going to break down the 10 ways AI can be used throughout the commercial real estate industry.
Understanding The Types of AI Available
When many people think of Artificial Intelligence they just think about AI in general. However they do not realize there are actually different types of AI. The two we are going to review is Generative AI and Predictive Analytics. Both of these are important in your technology stack to create a holistic solution.
Generative AI (GenAI)
Generative AI works via training deep learning models called (neural networks) based on massive datasets to understand patterns, structures, and relationships which helps to enable it to predict the next most likely elements after processing a user’s original prompt. From here the system will “generate” an an output such as (text, images, music, code etc etc that resembles the training data. This is a common reason why AI generated information all sounds the same.
In terms of commercial real estate it is good for some of the following items:
Summarizing, drafting, extracting information, and classification of text. Specifically, it is great for leases, reports, tenant comms, and workflow support.
Predictive/ML analytics
Predictive/ML analytics uses past data to train a model that learns patterns linking inputs to desired outcomes, then works to apply those learned patterns to new data to predict a value or probability that something may or may not take place.
This type of system works great to forecasts, detect anomalies, optimize schedules. It can be an excellent tool for pricing, risk, and maintenance needs.
The prerequisites That Ensures AI Works Properly
Many people are sadly under the assumption that AI does not work “well” this is due to macking simple mistakes that compound to produce poor outputs. Making AI work for you comes down to some key prerequisites such as the following:
- Clean Inputs
- Governorace
- Human Approval
- Consistent Output and testing
In the following sections we are going to break down each one to give you a greater understanding.
Clean Inputs
The first item on the list is “clean inputs” and for good reason without them you will get a poor output. As the saying goes “garbage in garbage out”. Clean input is about standardization, consistent naming and information flows. Ensure that any uploaded data follows proper data hygiene practices, is free of errors and contains all the necessary information you will need for AI to do its job properly.
Prompt Engineering
Once you have a standardized process for important company assets the next most important step is writing prompts that help AI understand what you are looking for it to do. If you provide vague information it will give you vague results. AI models have come a long way since being released to the world. To get a good result, provide as much information as necessary and allow the “magic” to take place.
Governance
Governance is an important part of the process to ensure you are able to stop people from influencing a model with bad information. This is where items such as access control, approved sources, versioning and most importantly audit trails come into the picture. Without the necessary guardrails in place you run the risk of “training” bad AI outputs.
Human Approval
Human approval is critical to the process of making sure that your desired output is happening as expected. You cannot hand over the reins to these systems expecting perfect results every time. This is especially important when it comes to legal, and financial decisions.
Consistent Output and Testing
If you have done all of the above steps you should be at a point where consistent output is becoming the norm. Depending on how many people inside the firm that are using AI someone will need to “keep-up” with the output results for whatever your initial prompt request was. Additionally, this is also the time to continually perform “tests” to maintain the models over time.
Diving Into 9 Practical Uses
Now that you have a general understanding of how the models work and how to make the systems produce the outcome you are looking to achieve. Lets dive into the practical uses of AI that will help advance the success of your firm.
Use Case 1: Lease abstraction and “red flag” detection
Lease abstraction is the process of extracting and summarizing crucial data from complex lease agreements such as commercial real estate contracts into a simple, standardized format, called a “lease abstract”. This summary highlights key details such as lease dates, rent payment schedules, renewal options, maintenance responsibilities, and critical deadlines.
This process used to take a significant amount of time to review documents at length. It can now be automated and performed faster using the help of AI. Additionally, it can also help your firm avoid missed details helping to identify any “red flags” your team may have missed during an initial review of documents. Abstracts are crucial to ensure CRE firms manage portfolios efficiently, avoid costly errors, and ensure compliance.
Use Case 2: Portfolio document search and instant reporting
Next on the list is efficient document searching. A few years ago you used to have to search documents, PDFs and email threads one by one to collect information in order to build reports. That is now a thing of the past with AI CRE firms now have the ability to search through documents and emails all in one centralized location. These include being able to search through leases, OMs, inspection reports and more.
Effectively saving you time, energy and money as time goes on.
Use Case 3: Leasing insights: lead scoring and tenant match-making
Gaining valuable insights into leasing activities, lead scoring and tenant match making can all be done the help of AI. A properly created system will be able to assess which leads are likely to convert based on various data signals. It can also quickly suggest “best-fit” suites based on questions answered by tenants. Further simplifying the process of day to day leasing operations and lead quality understanding over time.
Use Case 4: Market monitoring and competitive intelligence
Market monitoring and competitive intelligence is another important task that should be done regularly and with as much detail as possible. The problem? Your staff only has so much time in the day to review information that is required for mission success. With AI it is possible to continuously track comps, listings, leasing velocity signals as well as local activity to produce a “what’s changed this week? “ one pager that can assist your leasing and asset management team.
Additionally, for added benefit you can also tie-in relevant portfolio analytics, sort documents and standardize data.
Use Case 5: Underwriting copilot
Quality underwriting is one of the most important tasks you can perform as a CRE professional.
AI is transforming the commercial real estate (CRE) underwriting by automating time-consuming tasks, providing advanced risk analysis, while also enabling faster, more powerful data-backed decisions. It can also help you flag missing assumptions, run sensitive prompts and draft investment committee ( IC ) memos ( ensure you have a human review the output )
The shift in technological adoption enhances human expertise rather than replacing it, allowing underwriters to focus on strategic, high-value activities.
Use Case 6: Due diligence automation
The due diligence can take weeks or months to complete. AI systems can help the due diligence process in a the following ways:
Document Review & Data Extraction
AI can rapidly scans and extract key data points, such as clauses, renewal dates, and terms from thousands of contracts and documents helping to save hours of time manual review time. Always ensure that documents are reviewed for accuracy.
Risk Identification
Humans only have so much ability to dedicate time to a task. People get tired as time goes on. Mistakes can be made. A well built system can help spot hidden liabilities, financial irregularities, non-compliant clauses, or red flags in legal, financial, and operational data.
Enhance Analysis
Machine learning is excellent at identifying patterns and trends from large datasets, offering deeper insights into a target company’s health and market position in ways that would take a human an immense amount of time to analyze the same amount of data AI can do in minutes.
Automated Compliance
AI is getting smarter by the day and can help check for missing regulatory language such as GDPR, CCPA. It can also search for any inconsistencies across documents, to ensure adherence to internal quality standards.
Use Case 7: Predictive maintenance, smarter work order prioritization
Owing CRE property comes with maintenance requirements – a lot of them. Over time you will find certain things break other than others. Predictive maintenance is now possible with the help of these systems. They can be used to predict failures and recommend interventions be taken to minimize downtime.
Use Case 8: Building ops optimization
Building operations and optimization can be undertaken through the IoT (Internet of Things) whereby AI can identify abnormal patterns, find inefficiencies and take steps to optimize operations on a daily basis.
Use Case 9: Tenant experience automation
In commercial real estate tenants are the lifeblood of your business. Without them your firm would need to rely on brokerage transactions. Staying on top of tenants needs is a time consuming process. AI can help make that process faster and better. Your tenants will have complaints over time this is a natural part of owning property, they also will have recurring issues that you can help identified based on the number of request that
AI can be effective at identifying recurring issues, as well as summarizing your top complaints on a weekly / monthly basis so you can make improvements to operations and tenant experiences.
Risks and how to avoid the common failures
As great as Artificial intelligence is there are still some downfalls to it. These pitfalls can greatly derail your applied use. Some of these items are
Hallucinations
Hallucinations are where AI will make up information that does not exist. If not caught promptly these issues can cause major issues with your output. Luckily there is a solution which is to force citations and link back to any sources found. This will help you ensure you receive clean outputs.
Bad data
What you put into AI you will get out. If your data quality is poor then you can expect to get a poor outcome. The fix here is simple, apply standards to how you input and display information. The more context you provide the better your output will become over tome
Security
Security will play a major role in how you use AI at your firm. The best way to minimise issues is to restrict access to only the necessary parties required to complete a given task. Another good rule of thumb is to always ensure that your team does not upload any sensitive data prior to approval.
Conclusion
There are two types of CRE firms, those that embrace new technologies and those that do not. The ones that do are rewarded with smarter and soother operations, client satisfactions and better returns for their investment portfolios. AI is a tool that can be used to create powerful results if utilized property and with intent. Do not get left behind by your competition. Now is the time to act. If you need help building custom systems, reach out to the Cindtoro team for an evaluation. Start making faster decisions, fewer errors and scale out predictable operations – not just prettier looking emails.
We Recommend: