Beyond the AI Hype: Atlas Experts Answer Common Business Questions

Artificial intelligence is developing at a pace that can make even the most technology-savvy business leader feel like the finish line keeps moving.

Companies are hearing that AI can increase productivity, automate processes, unlock insights from their data and transform the way employees work. At the same time, executives are wrestling with much more practical questions:

• Where should we start?
• Which opportunities will actually deliver a return?
• How do we protect company data?
• What happens if the AI technology we invest in today is outdated six months from now?

Those are exactly the kinds of questions Atlas Professional Services’ growing AI Advisory Services team is helping businesses answer.

We recently introduced Christian Lopes, AI Process Consultant, and Nihar Maddina, Forward Deployment AI Engineer, who bring complementary business process and technical expertise to Atlas’s AI practice. In this Q&A post, we get their perspectives on some of the most timely AI questions facing executives today.

Q: Many companies know they should be doing something with AI but don’t know where to start. What should business leaders evaluate first?

Christian: Don’t get too hung up initially on which AI tool you’re going to use. Start by identifying the problem inside your business.

Look for something that’s high volume. Look at places where decisions can only be made by a handful of people or where employees are consistently experiencing difficulties. You want to identify your real pain points first.

Then make sure the data you would need is available and organized. AI models are only as useful as the data they can access.

For example, if you’re considering a knowledge-based search tool, where are your documents? Are your statements of work, contracts and policies organized and accessible? Or does one department have documents in one location while another department stores them somewhere else?

Once you understand where the process is breaking down and what information is available, you can begin looking at where AI can assist.

Nihar: Organized data and documented processes open a lot of doors. There’s a reason people say, “Data is gold.”

But it’s not enough simply to have a lot of data. You want accurate, useful data. There is a difference between good data and messy data. If you have multiple versions of the same document, for example, an AI system needs to know which information it should rely on.

Q: Where are you seeing AI deliver meaningful ROI today, and what separates a valuable use case from hype?

Christian: A lot of the best opportunities are the mundane tasks people repeat every day. Maybe someone is reconciling information or taking data from one spreadsheet and putting it somewhere else. Those repetitive tasks can be strong candidates.

AI is also very good at working through large amounts of information. Think about contracts. Instead of having someone read five historical contracts to determine how something was handled previously, AI can help summarize that information and identify potential discrepancies.

Where the hype comes in is the assumption that AI will do everything from start to finish without errors and completely remove the human component. That’s not how we’re approaching it. You still want a person validating information and making important decisions. The goal isn’t necessarily to “set it and forget it.”

Nihar: There are absolutely places where you may want AI to run most or all of a process. But there are critical points where human interaction still matters.

The creative or distinctly human part of someone’s job can remain while AI removes hours spent on mundane work. That’s especially relevant in areas with repeatable, patterned tasks. The question should be: What is meaningful to this business? If a repeatable process is consuming significant time and AI can improve it, that’s an opportunity worth exploring.

Q: What are some business processes executives might not immediately recognize as opportunities for AI?

Christian: Knowledge-based search is a great example and can be relatively straightforward.

Think about all the HR documents, statements of work, policies and documented business processes within an organization. Employees often know information exists, but they don’t know where to find it.

You could build a chatbot or an integration within a collaboration platform where an employee asks for a policy and gets the relevant information, along with the source. Instead of asking five people and eventually finding the one person who knows where the document lives, you may be able to get the answer in seconds.

Another opportunity is comparing historical documents.

Let’s say you’re creating a new statement of work. AI could help compare it with prior agreements and flag differences in language or pricing. Maybe the business charged $150 an hour for a particular service previously, but a new agreement lists $100 for essentially the same thing. AI can flag that discrepancy so a person can review it.

And then there are all those little tasks employees barely think about because they’ve become part of the job. If someone spends 45 minutes every day taking information from emails and putting it into a CRM, those minutes add up quickly over weeks and months.

Those are exactly the types of processes worth examining.

Nihar: Businesses should also look carefully at what data they’re collecting and what they’re not collecting.

Knowing what information you have can show you what’s possible today. Knowing what you don’t have can help you determine what you should begin collecting so you can analyze it and improve the process in the future.

Q: What’s one of the biggest mistakes businesses make when employees begin using AI on their own?

Christian: One of the biggest issues is “shadow AI.”

An employee in accounting might be using one AI tool while someone in marketing is using another, and business leaders may not even realize those tools are being used.

If there aren’t established policies or guardrails, employees may put sensitive company information into tools that haven’t been evaluated or approved for that type of data.

That’s why businesses need governance. Employees should understand which tools they can use, what information can be shared with them and how AI should be used appropriately.

Nihar: There’s another side to it as well. Some companies are already doing impressive things on their own. Employees are creating projects, workflows and integrations with AI, and they may get a large portion of the way toward what they want. But getting from an experiment to reliable automation can require another level of engineering.

That “last mile” is where an AI team can help. A prototype that works for one employee isn’t necessarily the same thing as a secure, repeatable business system that can be relied on across an organization.

Q: How can businesses establish AI governance without making the technology so difficult to use that employees work around the rules?

Nihar: One approach is to create a controlled enterprise environment rather than allowing employees to spread their usage across many different consumer AI tools.

Choose enterprise-grade solutions with appropriate administrative controls and data protections, then give employees a defined place to work. That gives the organization greater visibility and a more manageable point of control.

You also want to understand the provider’s policies around how business data is stored, protected and used.

Christian: Start small and make sure everyone understands the day-to-day expectations. You don’t have to roll out everything at once. Put the right protections in place, demonstrate value and expand from there.

Governance doesn’t have to mean putting AI behind so many locked doors that nobody bothers opening them. The objective is responsible use.

Q: AI is changing so quickly. How can a business invest today without worrying that its solution will be obsolete tomorrow?

Nihar: There will always be something newer. That’s probably not going to change. The point isn’t to compete with the newest AI models. The important thing for a business is to have its data intact, organized and accurate.

Our role is to use the best available tools to provide a solution designed around that client. We’re creating a bridge that helps bring these capabilities into small and midsized businesses in a practical way.

Christian: We try to be tool-agnostic. The tool is the tool. What’s more important is your process, your workflow and your data.

If the use case is sound, the workflow is sound and the data is sound, we can change the technology we’re using much more easily as better options become available. Whatever tool is best for the job is the one we want to use.

That’s an important distinction for executives evaluating AI solutions for businesses. The long-term investment isn’t simply in today’s model or application. It’s in building better processes, stronger data foundations and a clear understanding of what the business is trying to accomplish.

Q: If a CEO wants the company to be meaningfully ahead in its use of AI 12 months from now, what should the organization have accomplished?

Christian: Have your strategy documented. What tools are approved? What tools aren’t? What use cases are you pursuing? Employees should understand what they should and shouldn’t be using.

You also need to define value and track it. Is the value time saved? Is it money? Is it increased accuracy?

Then look at adoption. Did employees use the solution for two weeks and stop or has it become something they can’t imagine working without?

Your strategy should be fluid, because the technology will change. But you still need a strategy and you need to measure whether you’re accomplishing what you set out to do.

And, once again, make sure the data is sound and secure.

Nihar: Your team’s goals also need to be aligned.

Being as “AI-forward” as possible isn’t automatically a good thing. The point isn’t to use the most AI. The point is to make the business more effective.

The human processes are still what everything revolves around. AI is there to optimize them and make them more efficient, not simply replace the human component.

Q: What’s your favorite part of the AI Advisory work you’re doing at Atlas?

Christian: It’s when someone starts out a little hesitant and wonders whether AI is really going to work for them. Then they see the value and it turns into, “I don’t think I could live without this anymore.”

There’s a lot of skepticism, because AI has become such a buzzword. But when people see that it can get repetitive tasks out of their way and give them more time to focus on what they actually get paid to do, that’s rewarding.

Nihar: For me, it’s showing someone that AI can do something they hadn’t imagined.

There are use cases that go far beyond chatbots. Take manufacturing, for example. Computer vision can be used to identify patterns and potential defects in products during production.

When a client realizes, “I didn’t know AI could be used for that,” that’s exciting. You’re introducing them to a completely new way of approaching a problem.

From AI Curiosity to an AI Strategy

Perhaps the clearest takeaway from Christian and Nihar is that becoming an AI-enabled business doesn’t start with finding the flashiest new tool. It starts with knowing your business.

Which processes consume employees’ time? Where does information get stuck? What tasks are repetitive? Is your data organized enough to support automation? What should employees be allowed to do with AI? Most importantly, what outcome would make an AI investment worthwhile?

Those questions are at the center of Atlas’s approach to AI solutions for businesses. Its AI Advisory Services are designed to help organizations discover high-value opportunities, build a strategy around them and move from experimentation to secure, measurable implementation.

Wondering where AI could create meaningful value in your business? Atlas can help you evaluate your workflows, identify high-impact opportunities and develop an AI strategy grounded in measurable business goals. Contact Atlas Professional Services to start the conversation.