The construction industry faces a looming challenge — by 2030, the U.S. Bureau of Labor Statistics projects a significant shortage of skilled workers. With workforce exits by Baby Boomers and Gen X coupled with increased demand across industries, construction companies must act now to remain competitive in an increasingly tight labor market.¹

To address this issue, forward-thinking leaders are turning to technology and artificial intelligence (AI) as potential solutions.

In this conversation (which was initially recorded and transcribed with the assistance of AI), Joseph Harper, CFO of Kelley Brothers Roofing, and Rishi Srivastava, CEO of Beiing Human, discuss how AI-driven solutions can enhance administrative efficiency, streamline processes, and ultimately create a more attractive and engaging workplace for future talent.

Joe: Rishi, thank you for speaking with me to better understand how we can proactively adopt technology. I can’t help but feel nervous about the year 2030 — it’s clear that we will be vying for construction administrators in a tougher labor market. To manage this and avoid a crisis, planning needs to begin now.

The construction industry already competes with a wide range of industries for candidates. So how can leaders better plan to make their workplace more attractive to candidates while ensuring they retain their current workers? How can technology and AI be effectively used to not only mitigate workforce shortages, but also enhance employee engagement by increasing their value and sense of purpose?

Rishi: Those are great questions. Over the past few years, we’ve witnessed rapid advancements in AI, and it’s poised to make a major impact across industries, especially in administrative tasks like accounts payable (A/P), data entry, and validation. Phone automation is another area where AI can add value. Many administrative and marketing tasks can now be easily assisted by AI.

For example, AI-driven systems can handle data entry for A/P, while a human focuses on validation, freeing them to perform more strategic tasks. 
AI can also screen incoming calls, prioritize them, and determine which ones require immediate human attention, such as high-priority customer inquiries.

These are all critical tasks in the construction industry that can be transitioned to AI with the proper algorithms.

Joe: I have some understanding of algorithms, but in simple terms, what really is AI?

Rishi: In simple terms, AI refers to algorithms built on neural networks rather than simple mathematical models. The human brain is a complex neural network system, and an AI model mimics part of this structure through specialized training.

Another way to think about AI models compared to other models is that the types of problems they solve would be extremely difficult to handle using rule-based or purely mathematical approaches.

For example, determining the total value on a PDF invoice cannot be solved with high accuracy by writing thousands of rules. However, an AI model is perfectly suited to solve this type of problem.

Joe: I really like that answer. As we move toward the 2030s, I know companies are going to have to be more attractive to candidates — ours included. It is not just about attractive wages and benefits, but candidates and employees need to feel engaged in the work they are doing. Part of that engagement is knowing that their role provides a true value-add to the overall production.

A well-functioning AI-assisted call answering process will free a properly trained and motivated employee to spend the time addressing customer needs and concerns. This employee would have more opportunities to personally engage the customer. Employees that are provided with the opportunity to engage customers are more likely to understand how they add value to the overall organization.

So, let’s start with AI’s role in phone call routing. Is that a chatbot?

Rishi: No, it’s not really a chatbot.

There will be something called interactive phone technology. In my view, the biggest thing AI can do is effectively screen calls and route them to the 
proper place.

Designing a properly functioning chatbot involves so much more than pure AI models. There is an element of solid intent detection that requires deep business logic to build a properly functioning chatbot.

Regardless of the current hype of large language models (LLMs), installing chatbots in many situations won’t provide the return on investment needed and can cause customer frustration.

Joe: We don’t want to create customer frustration. It sounds like what we need to discuss is AI-driven call screening and routing.

Ideally, urgent calls that require immediate attention should go directly to an employee’s phone — these could be calls from customers, certain suppliers, or other team members. If the call is less urgent, such as a cold call or a sales inquiry from a non-vendor, then it would be routed to voicemail. Is that possible with AI?

To give an example, our company has an estimating department whose main goal is sales. If a call comes in from an existing customer, such as a GC, asking for an estimator, then I want that call to be routed to the responsible person’s phone.

But a call from a supplier who is simply letting everyone know there are updates to prices is a lower priority. That call would preferably get routed to the estimator’s voicemail. Is this level of prioritization something AI can help with?

Rishi: Definitely. This can be achieved by using AI to analyze the voiceprint of the caller. It can map that voiceprint to a person.

If the caller is an existing customer listed in your customer relationship management (CRM) system or enterprise resource planning (ERP) system. For example, if a GC that your estimating department is actively working with calls, AI can route it directly to the estimator’s phone.

On the other hand, if it’s a sales call from an unknown contact not mapped to any of your current prospects, then AI can route it to voicemail.