Jay Bhatty is the CEO and founder of NatGasHub.com.
Over the last two decades, software as a service (SaaS) has changed how we do business. Now we’re on the edge of another major transformation as agentic software gains traction in our personal and professional lives.
When you use a traditional SaaS application, you interact with what’s called a CRUD database—a framework that enables you to create, request, update and delete data. As a user, you don’t need to understand how the complex database works in the background. You only need to know how to manage content within the app, whether you’re watching your favorite show on a streaming service or revising a lead in your CRM.
Agentic software will simplify these interactions. Each application you use has its own foundational database that you currently have to query individually.
Agentic Software In Action
For example, if you’re planning a trip, you first need to research and book your air travel on one application. Then you can compare prices and book your hotel room or vacation rental on another application. Once you have your airfare and lodging in place, you can begin building your itinerary—looking at travel magazines for sightseeing recommendations or booking dinners on a restaurant reservation app. But you’re bound by the time constraints of your everyday life and the upcoming trip. Airfare prices may drop early on a Tuesday morning when you’re not available to buy tickets, and you could miss your window for the best prices.
With agentic software, you will soon be able to set up an AI agent to do each of these steps for you in the most optimized way. You can give your agent logins to each relevant account, along with specific instructions on the task you want completed. For instance: “If the price on this flight drops below $300, buy it.” If the best price pops up in the middle of the night, no problem, your AI agent is on it. Then you can give it further instructions on organizing the rest of your trip: “After you’ve booked my flight, find me a hotel room within this budget range in this neighborhood. Then reserve a table at an Italian restaurant downtown for Saturday night, compile a list of sightseeing highlights for the city and email me my itinerary.”
Generative AI assistants, such as ChatGPT and Copilot, are already managing many of these tasks. A GenAI chatbot can immediately generate a three-day sightseeing itinerary in Los Angeles for you, based on places you like. Agentive software will build on these existing capabilities. Instead of having to log in to multiple accounts and manually complete multiple tasks, you can assign one AI agent to move across different applications and execute functions within set parameters.
My company is starting to incorporate AI agents into all of our datasets. Our clients in the oil and gas industry have historically had to visit multiple pipeline websites every day to aggregate, synthesize, cleanse and standardize data. AI agents can now follow a precise set of instructions to automate these tedious tasks and give employees back time to do more interesting strategic work. We plan to expand this service to offer individualized agents that digest and analyze data based on each client’s specific needs.
I predict that agentic software will disrupt and replace traditional SaaS applications within the next few years. Soon, you won’t be interacting with just one AI agent but swarms of agents. You’ll have one agent for work tasks, one for chores and errands, one for travel and leisure planning and so on.
Strategies To Prepare For An Agentic Future
Agentic software is advancing quickly. Here are my top three recommendations for getting ready for this shift.
1. Recognize The Trend
It’s important to understand that we are on the verge of a major software transformation. Whether you build or use SaaS models, your business—and how your employees and customers interact with software—is going to change. If you get stuck in the old way of doing business, your business might be at risk.
Look at the positive aspects of agentic software. We all want simplicity and ease. Unless you’re a software developer, you don’t want to write a SQL or Python query to extract data. You can save an enormous amount of time by giving an AI agent verbal or written instructions and letting it do the rest.
2. Build Guardrails
As AI agents become ubiquitous online, companies need to build safety mechanisms into their actions. Intellectual property will be even more valuable, and paywalls are likely to become widespread.
Educate your users on best practices for setting parameters and implementing AI agents. Help them shift their mindset and get comfortable using an unfamiliar technology while still protecting their sensitive data. They likely already have their credit card information saved on shopping websites or have automatic bill payment set up for their utilities. Once they see that AI agents can effectively manage buy-sell decisions on a small scale, they’ll feel more at ease moving on to bigger ticket items.
3. Rethink How You Leverage Human Labor
A question I hear a lot on panels and in presentations is: If AI agents are going to do all of this work, then what are humans going to do? The answer isn’t that humans will end up jobless while machines do all the work. AI agents will take over the tasks that are boring, repetitive and time-consuming—and what employees end up doing with that time is full of possibilities.
Think about all the time you would free up if you didn’t have to spend time scraping data from websites or trying to find the best rate on a flight. We will open up hours of time for people to work on more mentally challenging tasks—from expanding quantum computing to building faster airplanes. There are limitless conquests that humans can make.
AI agents aren’t here to take human jobs; they’re here to give us time to tackle the higher-order problems that agents aren’t able to solve.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
