Jay Bhatty is the CEO and founder of NatGasHub.com.
For companies in the technology industry, tech is a revenue generator. For companies in every other industry, tech is a cost center. You aren’t making the software or providing the service; you are buying it from a third party and paying to maintain it.
But as AI becomes more ubiquitous, successful companies outside the tech industry are now discovering how to convert technology from an expense into a competitive advantage. Here are three key lessons I’ve learned from the energy sector that you can apply to your organization, no matter what industry you’re in.
1. Intelligent Document Processing
Every company has to manage incoming and outgoing invoices with customers, vendors, suppliers and other parties it does business with. And every company’s invoice has its own format and design. There is no standardization. Even within the same organization, two departments can have two completely different invoices.
Document processing can be extremely labor-intensive, but with the advent of large language models (LLMs), you no longer need to dedicate hours of internal or outsourced human time to this task. If you feed an LLM a PDF invoice, composed of dozens or hundreds of pages, it can digest and understand all of the content in that document. It acts as a digital agent that you can interact with, asking it questions and giving it specific instructions to standardize, approve and pay that invoice.
I predict that LLMs will soon become a commodity that any business can afford. In the same way that you can buy a laptop produced by Dell, HP, Google or Microsoft, you’ll be able to buy an LLM that can process and standardize your data better, faster and cheaper than a person can. Once you make back your initial investment, the employee time that you save goes straight to your bottom line.
LLMs learn and improve over time, but they can’t operate without quality data and human oversight. My company builds AI software specific to the energy industry. The LLM learns the details related to natural gas or coal delivery, for example, and improves its speed and accuracy with each invoice. But we have to limit the amount of data it can access and train it not to use the generative component of AI for this particular task. There is nothing new to generate from an invoice. We don’t want the LLM to hallucinate and create a charge that doesn’t exist, so we train it with specific commands, such as: Only look at these existing charges, and do not create any new line items.
2. Robotic Process Automation
People in every industry have concerns about AI eliminating jobs. But I’ve seen in the energy industry that AI can be a job-enhancer, not a job-killer. AI can take over the boring, repetitive and error-prone tasks that humans are happy to give up.
Robotic process automation (RPA) is a form of digital labor. You train an AI bot to do simple, recurring tasks that require no decision-making, such as logging in and out of an account, clicking buttons and copying and pasting data.
For example, if an employee goes to a news website every morning, downloads market prices, creates a chart from the data and emails it to 30 co-workers, this is an ideal job to outsource to RPA. Your employee gets back a block of time each day that they can spend on something more revenue-generating and mentally stimulating, helping you keep team morale high.
Unless your company builds software, I don’t recommend building RPA bots internally. You can easily transfer labor cost savings to technology expenses. It’s important to find a software company that specializes in building and running bots for clients in your industry. RPA is similar to managing people. You need to make sure that bots show up and do their jobs to completion every day.
3. Automated Asset Platforms
Utility companies have found an efficient way to manage stranded assets. If there is an unseasonably warm day in the middle of winter, for example, the local utility doesn’t need to pay to transport or store gas or electricity.
Instead of charging customers for these stranded assets, they can sublet the pipeline space or electricity wires they don’t need in downtime on a secondary energy market. Or if they have an unexpected energy shortage, they can go to the market and buy it from another utility. It’s similar to subletting your apartment. If you live in a big city and you’ll be traveling for the summer, you don’t want to pay rent for three months on an unoccupied apartment.
Automation now enables easier decision-making for this type of secondary asset market. Energy traders can use automated platforms to analyze historical data—from the last five winters, for instance—to determine accurate pricing.
This type of technology already exists in industries including energy, real estate and entertainment—but it can be adopted in other sectors. What stranded assets are a current cost for your business that you could resell to turn a profit? Are there existing, trusted marketplaces in your industry? You want to go to a platform that already has a deep pool of buyers and sellers versus trying to do it yourself. For instance, my company provides data to a global commodity exchange where traders go solely to buy and sell energy. They know they will find exactly what they want to buy and sell on this platform.
I encourage all business leaders who are not in the technology industry to stop thinking about tech as a cost. Look at it as an investment that has an ROI. Flip the paradigm and view the problem from a different angle. Instead of trying to figure out how to cut your tech expenses, explore how tech can become a revenue generator.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
