To simply put, you will probably no longer have to learn SaaS applications to automate workflows. An ‘intern’, sorry, a ‘professional’ in the form of AI Agent will take care of the same work. No more learning SaaS apps, and this trend is really bad for legacy applications out there.
Also, this ‘idea’ of ‘AI Agents to kill SaaS’ is not a ‘too soon’ scenario unlike AGI. This prediction started to surface in December 2024. As of February 2025 of writing this article, I am already seeing AI agent tutorials all over YouTube!
“So Moore said: computers would double in capacity every 18 months.
And AI came along and said, hold my beer. AI models are doubling every six months. And this growth is partly what’s causing all this excitement around AI.”– Dharmesh Shah, CEO at HubSpot | Source: INBOUND 2024
So I dug deep into this debate of ‘Will AI Agent kill SaaS?’ and now there is more than enough picture drawn for me to summarize and extrapolate on. I have shared ideas from AI and entrepreneur experts that you can ruminate or capitalize on.
But, what is even ‘SaaS’?
Let me give you a simple conceptual breakdown of how ‘software’ has evolved over the years.
Remember the times when one used to ‘download’ software?
A very common example is of ‘Microsoft Office’. One had to separately buy the MS Office ‘Computer Disk (CD)’ and INSTALL it into your computer or laptop. This was the time when software was purchased as a LICENSE and installed locally. As a result, companies had to often set up data storage centers to make these software work. Also, programs were built for specific hardware and could not easily interact with other systems.
The ‘internet’ technologies spread rapidly like wildfire in the early 2000s. ‘Cloud computing’ also was a promising new technology that democratized storage. A new model emerged from this expansion: Software as a Service (SaaS). SaaS transformed the traditional software delivery model by offering:
- Accessibility: Users can access SaaS applications from anywhere with an internet connection, fostering a more connected and agile workforce.
- Subscription-Based Pricing: Instead of a one-time buy, users subscribe to software and pay a recurring fee.
- Centralized Management: Software is hosted in the cloud, meaning updates, security, and maintenance are handled centrally.
- Scalability and Flexibility: Companies can scale their usage up or down based on demand without investing in physical infrastructure.
Thus, with SaaS, one reduced the burden of local IT management while enjoying continuous improvements over time. SaaS quickly became the standard for delivering enterprise applications across all industries like CRMs, project management, design, etc. Many billion dollar worth companies were born that one can access from anywhere anytime.
This includes giants like WordPress, Figma, Shopify, Salesforce, SAP, Slack, Spotify, Zoho, CleverTap, and many more pop up even today.
We have come a long way too when it comes to ease of SaaS development. In the past, people spent huge dollars to create software. Today, we have no-code tools to spin off SaaS in a matter of hours.
Image Source: inVerita
Strangely, moving into the ‘cloud’ from ‘local storage’ to host software is still an ongoing trend. Many offline-heavy industries like manufacturing or local shops continue to rapidly digitize thanks to SaaS massively bringing down digitization costs.
When all is well with SaaS, why are AI Agents relevant?
AI Agents are smart, and we humans love doing less.
With SaaS, you have to learn software, go through its documentations, figure out automation, manage workflows, and pay up monthly/annually.
AI Agents, or Agentic AI, unlike SaaS, are not just passive applications. They are designed to operate autonomously. AI Agents are capable of making decisions and executing tasks with minimal human intervention. Thus, we can do away with above-mentioned SaaS management tasks.
“What an agent is? It is an AI that actually completes a task. You ask it to do something and it just doesn’t just tell you how to do it – it does it for you.”
– Konstantine Buhler of Sequoia Capital, in conversation with Bloomberg Technology | Source: Why 2025 Will Be The Year of AI Agents
In other words, these agents aren’t just fancy algorithms. They’re digital workers capable of handling complex, multi-step tasks. They integrate various data sources and perform actions on behalf of users. They can be simple or incredibly sophisticated, but all share three core traits:
- Use of AI Models: Without AI, there’s no agent—only automation.
- Tool Integration: Agents can connect with platforms like HubSpot’s Smart CRM, among others, to perform tasks.
- Memory: Much like us, agents can “remember” past actions or data points, ensuring continuity in task management.
AI Agents will bring huge productivity boost for organizations across industries and operating levels thanks to below key features:
AI Agents are autonomous
AI agents can perform tasks end-to-end without constant human intervention.
For example, in a customer service context, an AI agent can analyze customer queries. It can access relevant data across various platforms and respond in real time. A customer service representative does not need to step in. This kind of end-to-end autonomy increases efficiency. It also reduces costs because fewer human resources are needed for routine tasks.
AI Agents adapt and learn
Today, we have new, advanced, and cheaper AI models released every week. They will power these AI agents to continuously learn and adapt to new data to refine their processes dynamically.
Businesses do not need to wait for a software update or a new feature release from a SaaS vendor. They can adapt their AI agents in real time to meet new challenges. This level of responsiveness is increasingly valuable in a fast-paced, competitive environment.
Consider again the above customer service example. This means that over time, an AI agent can become more effective at predicting customer needs. It can improve workflows and even foreseeing potential problems before they arise.
AI Agents integrate across systems
Traditional SaaS tools operate in siloed environments. AI agents are designed to interact with multiple systems simultaneously. They can access databases seamlessly and bridge disparate data sources effectively. AI agents can bridge these gaps by interfacing with various databases, SaaS tools, and enterprise systems.
This cross-system integration enables them to make holistic decisions that take into account a broader context. This is something that isolated SaaS applications are ill-equipped to handle.
AI Agents will get cheaper
Traditional SaaS platforms typically charge on a per-user or per-seat basis, often leading to escalating costs as businesses scale.
In contrast, AI agents operate on a task-based or consumption-based pricing model.
This pricing structure can be far more cost-effective, particularly for large enterprises that execute many repetitive tasks.
However, one may argue – Generative AI hallucinates and is not reliable with facts. But as Sam Altman said – “This is the dumbest AI will ever be.”
Here’s how YCombinator thinks about where AI Agents are today in comparison to what SaaS was decades back:
I think Salesforce is probably like the first true SaaS company. I remember Mark Benov coming to speak at YC. He tells the story of how, very early on, people didn’t believe you could build sophisticated Enterprise applications over the cloud. They also doubted such applications could be built via SaaS. There was like a perception issue. It was like no, you buy your box software and that’s like the real software. That you run the way we always do it. It was quite contrarian because the early web app sucked. You had to be a Visionary like Paul Graham and understand that the browser was going to keep getting better. That eventually it will be good. (This is) quite reminiscent of today where it’s like – no you won’t be able to build like sophisticated Enterprise applications that use these LLM or AI tools because they hallucinate. Or, they’re not perfect or they kind of like just toys – but yeah that’s like the early SaaS story.”
– Panel discussion on ‘Vertical AI Agents Could Be 10X Bigger Than SaaS‘ by YCombinator
What will AI Agents ‘change’ about software development?
SaaS provided a reliable and accessible way to deploy applications. It was often built on fixed business logic, the so-called CRUD model: create, read, update, delete.
Today, Agentic AI is poised to replace much of this hardcoded logic with dynamic, autonomous decision-making. Instead of relying on pre-coded instructions, these agents learn and adapt to the specific needs of an organization. This helps deliver a level of efficiency and personalization that traditional SaaS applications simply cannot match.
That’s the change, explained further by Satya Nadella:
“Whenever there’s been a real platform shift the core application architectures have changed. I think what’ll happen is these CRUD, I mean, SAAS applications are a CRUD database with a lot of business logic. So the CRUD database will then get orchestrated outside of the business logic tier of just the SAAS application is what I mean is going to happen. Like, right now in my own use case I go to co-pilot I say at sales which is actually touching Dynamic CRM brings back whatever the account information, then it brings back information from Office 365, I put it into pages I share it with people the entire workflow. I mean everybody talks about their CRM database but nobody uses it because you know when was the last time I logged into CRM? Never. Except now, I’m every day querying my CRM database because it’s so much easier – it’s one agent away and it’s working with all the other agents. So, that is what’s going to be the change.”
– Satya Nadella, CEO at Microsoft, in conversation with Varun Mayya, source: Applied AI Tools
I have further covered two key ways in which AI Agents will change how we develop software – Transforming Software Design with AI Agents
Going from Vertical SaaS to Vertical AI Agents
Before diving into Vertical AI Agents, let us understand what is ‘Vertical SaaS’?
Vertical SaaS refers to cloud-based software solutions that are designed specifically for a particular industry, niche market, or problem statement. This is unlike horizontal SaaS, which targets common business functions across many industries (like CRM, ERP, or HR systems). Vertical SaaS accommodates specific workflows and regulatory requirements.
For example, instead of a general documentation software, you can design specialized legal practice management software. This legal vertical SaaS will include document automation, case management, and time tracking. This would be more helpful for law firms to manage their practices effectively than general documentation software.
Adopting vertical SaaS means businesses can avoid the pitfalls of trying to retrofit generic software to meet their needs. Instead, they can use platforms that are already designed with their unique requirements in mind. This reduces the need for expensive, time-consuming customizations, thus enabling a more seamless digital transformation.
However – building a vertical SaaS is not easy.
Here’s YCombinator panel explaining why major technology companies like Google or Amazon did not clone successful vertical B2B SaaS businesses. They had the money, scale, and resources – but they still did not choose to:
“It’s just too hard to do that. Many things as a company like each B2B SaaS company really requires like the people who are running the product the business to be extremely deep in one domain and care very deeply about a lot of really obscure issues. Take like Gusto for example – why didn’t Google build a Gusto competitor? Well, there’s no one to Google who really understands payroll and has the patience to deal with all the nuances of all these like stupid payroll regulations. It’s just not worth it for them. It’s easier for them to just focus on like a few really huge categories in the B2B SaaS world.”
– Panel discussion on ‘Vertical AI Agents Could Be 10X Bigger Than SaaS‘ by YCombinator
Thus, opting for a vertical SaaS business provides a solid ‘moat’ or ‘unfair advantage’ against established businesses.
Now, back this up with an AI agent, that is, building an AI agent in vertical SaaS domain.
What are Vertical AI Agents?
Vertical AI agents are specialized software solutions that combine advanced large language models (LLMs) with domain-specific expertise. They are designed to automate repetitive, administrative tasks within a specific use case. This can be in any industry – be it quality assurance (QA), recruiting, customer support, or even voice-activated services.
We are now witnessing the emergence of full-on vertical AI agents.
By focusing on niche functions, Vertical AI agents will improve efficiency and create entirely new categories of enterprise software. Founders are encouraged to explore “boring” and repetitive administrative tasks. These areas may have been overlooked in the past. Now they are ripe for disruption by AI!
Looking ahead, Vertical AI Agents will grow into intelligent systems designed to replace entire teams and functions within enterprises.
When I first read about this future of work prediction, I was quickly reminded of my struggles. I still find it challenging to get quotes from audio and video files. I still do it manually, and did for this blog too!
But then, that is how web apps were initially – clunky and limited. But as technology evolved, so did the sophistication of these platforms.
Today’s vertical AI agents follow a comparable trajectory. They will evolve from simple text-based applications that hallucinate. These systems will handle complex tasks and integrate deeply into enterprise workflows – patience is the key!
The idea of this blog is to know about the rise of Vertical AI Agents. Keep an eye on the latest trends and you can either make one or use them for best results.
Vertical AI Agents: a new business model opportunity
The economic impact of vertical AI agents could be staggering.
Nearly 40% of venture capital dollars over the last 20 years were funneled into SaaS companies. This led to the creation of more than 300 SaaS unicorns.
Today, many industry leaders increasingly believe that vertical AI agents will disrupt existing SaaS solutions. They will also replace significant portions of the workforce presently dedicated to routine tasks. AI systems will take over tasks like payroll, customer support, and even internal communications. As a result, the cost structure for companies will change dramatically.
“There’s a sense that as Revenue scales the number of people you have to hire scales with it. So when you look at unicorns even in today’s YC portfolio, it’s quite routine to see a company that reached a hundred or $200 million a year in Revenue. But they have like 500 to 2,000 employees already. The kind of advice I would give last year or two years ago in the past you might say – let me find the absolute smartest person in all of these other parts of the org customer success or sales.
I’m starting to sense that the meta is shifting a little bit. You actually might want to hire more really good software engineers who understand large language models. Who can actually automate the specific things that you need that are the bottlenecks to your growth. So it might result in a very subtle but significant change in the way startups grow their businesses post product Market fit. It means that I’m going to build LLM systems that bring down my costs that cause me not to have to hire a thousand people.”
– Panel discussion on ‘Vertical AI Agents Could Be 10X Bigger Than SaaS‘ by YCombinator
The dream of running a unicorn company with a lean team may soon become a reality.
The opportunity with Vertical AI Agents is to help companies become lean.
How big is the Vertical AI Agent opportunity?
“Here’s my pitch for 300 vertical AI agent unicorns. Literally every company that is a SaaS unicorn you could imagine there’s a vertical AI Agent unicorn equivalent. It is like some new universe. Most of these SaaS unicorns beforehand there were some box software company that was making the same thing that got disrupted by a SaaS company. You could easily imagine the same thing happening again.”
– Panel discussion on ‘Vertical AI Agents Could Be 10X Bigger Than SaaS‘ by YCombinator
While we will have robots doing the dangerous jobs. With vertical AI Agents, it will be the boring jobs that no one wants to do. That’s going to be a lot of work to build and automate for!
How to explore opportunities in Vertical AI Agents?
We can understand Vertical AI Agents by below two concepts on how existing businesses are implementing and adopting it:
Offer hyper personalized solutions using Vertical AI Agents
The opportunity is big, one must focus on going extremely niche.
Niche is not just doing down on one industry.
It is down to the point of doing one task exceptionally well for that one client.
Hyper-personalize your solution to help automate customers with Vertical AI Agents.
Here’s YCombinator explaining the case for hyper-verticalization using example of GigML and Zepto (an Indian instant delivery startup).
“I could see that being a case of hyper specialization or hyper verticalization.
Maybe eventually there could be a single general purpose customer support agent software company. But that will be like a eighth or ninth inning, and we’re literally in the first inning.
So, instead, you’re going to have companies like GigaML. It’s doing 30,000 tickets every single day for Zepto. It is replacing a team of a thousand people. But it’s very specific and it’s not a general purpose. It’s 10,000 test cases in a very detailed set that is basically just for Zepto and things that Zepto does.
But if you are any of the other marketplace companies, you’re probably going to use it. It’s a very well-defined marketplace. You know, instant delivery Marketplace.”
– Panel discussion on ‘Vertical AI Agents Could Be 10X Bigger Than SaaS‘ by YCombinator
Hence, you can focus on a hyper-niche problem statement. Based on practical results, you can scale it to cover more use cases or scale to higher niches. In the example shared, one can scale from a Zepto-specific problem. This expands to solving for ‘Instant Delivery’ marketplaces. Eventually, it extends to marketplaces as a whole, and beyond.
What problem statement comes to your mind to build an AI Agent for?
For me, I would like to have quotes taken out with precision from audio or video files. Let me know yours in the comments!
Building an Agent Builder to help others build AI Agents
If going hyper-vertical sounds too much work – why not build a platform to ‘build AI Agents’?
We will understand this with the example of one of the largest SaaS companies – HubSpot.
In its recent INBOUND 2024 conference, Dharmesh shared his excitement for AI Agents space. He also talked about what HubSpot is doing in it.
Understanding HubSpot’s steps in AI Agents space will give an insight about:
- How existing SaaS can navigate Agentic AI revolution to adapt.
- How to conceptualize AI Agent Builder platforms.
- Gives a sneak peek into how AI Agents would look like in professional domain.
Dharmesh released ‘Agent.AI‘ – the only professional network for AI Agents.
You can explore existing AI Agents in the marketplace, build AI agents using its Agent Builder, or request AI Agent.
[Altogether,] there have been 3,420 agents built on the platform – of which 302 have been publicly shared.
– Dharmesh Shah, CEO at HubSpot, in conversation with CX Today
Dharmesh predicts that teams of future will be ‘hybrid’ – consisting of humans and AI agents.
“Now what if I told you agents can use other agents? I’m telling you, agents can use other agents. We can use existing agents as building blocks for building new, more powerful agents. It’s like agent composition. I haven’t had this much fun since I played with Lego.”
– Dharmesh Shah, CEO at HubSpot | Source: INBOUND 2024
When I explore Agent AI, the website design looks like ‘LinkedIn’. Each AI Agent has its own profile, ‘Add to team’ button, reviews, tasks completed, and more.
To me, the platform seemed like a set of ‘mini tools’ that one can use for highly specific tasks. Since the platform is free, I do use it a lot and would recommend you to check it out!
For example, let’s explore the ‘Company Research Agent’ developed by Dharmesh himself and research ‘HubSpot’:
It shared to me a very detailed report that covered everything about HubSpot. This includes, funding status, website traffic, competitors, news, social statistics, potential business challenges, PPC analysis, and so much more! You can even ask questions and it will share answers based on this report.
You can check the report here to explore Agent AI’s interface without signing up – Agent AI HubSpot Company Report
The Agent AI’s marketplace is separate from HubSpot as of today, but they plan to integrate it in the future. They already have 258,000 users. Agentic AI for small and medium businesses is a key focus for investment in 2025 for HubSpot Ventures.
“Our vision for the future is: there’s an agent for that.”
– Dharmesh Shah, CEO at HubSpot | Source: INBOUND 2024
Not just HubSpot, even Salesforce is working on ‘Agentforce’ – its own AI Agent Builder.
If you are an existing SaaS thinking ‘what’s next’? HubSpot and Salesforce provide a very good sneak peek.
That is, the rise of AI Agent Builders as a market.
Will you enter it? – let us know in the comments!
I am not an entrepreneur – what AI Agents can do for a mere mortal end user?
Satya Nadella thinks that you will build personal AI Agents like how you create spreadsheets and powerpoint slides today.
“When you’re hiring people, I think in the future you’ll now hire people plus their workflows. It’s like when you hire a data analyst you hire them and their spreadsheets. That’s kind of what it is right – so, AI agents are going to, I think two years from now, we’re going to say, ‘yeah I build them like all day like I build docs and spreadsheets and I think that’s going I come with a basket of them.”
– Satya Nadella, CEO at Microsoft, in conversation with Varun Mayya | Source: Generative AI quotes on future of work
You may need not purchase expensive SaaS products – but invest in platforms that help automate by building AI Agents. Just like you stopped writing because of MS Office or Google Docs.
“I predict that many of you will be building agents this year and not just the introverted developers like me. Anyone can build an AI agent. You can build an agent. Yes You.
Remember, you don’t have to be a coder to be an agent builder. You just have to be curious. All it takes is a few easy clicks.”
– Dharmesh Shah, CEO at HubSpot | Source: INBOUND 2024
Here’s how the HubSpot’s AI Agent Builder looks like from my dashboard. This is the edit screen for the ‘Chat with your file’ template. I can change it to include custom actions and triggers. Though a lot of data sync, share, and capture is linked to HubSpot CRM, of course.
Nevertheless, no-code builders like me are going to have a lot of fun with AI Agent Builders.
AI Agent Builders are only going to get more mainstream. Hence, it is of critical importance to learn how to build or use AI Agents.
I will be covering more about this topic, subscribe to AppliedAI Tools to learn more about AI Agents:
Note: There will be a transition period of convergence between SaaS and Agentic AI
Coming back to the main topic – will Agentic AI replace SaaS?
It is tempting to view the rise of AI agents as a binary replacement of SaaS.
However, history teaches us that technological transitions are rarely that simple.
Instead of one model completely eradicating the other, what is more likely is a convergence of technologies. In the coming years, SaaS platforms might embed AI agents directly into their solutions. This approach offers a hybrid model that leverages the strengths of both.
This convergence would allow companies to benefit from the stability and familiarity of traditional SaaS. They could also take advantage of the dynamic, adaptive capabilities of AI agents.
For example, a CRM platform might integrate an AI agent that autonomously manages follow-ups. It might also handle lead prioritization. However, users can still intervene when necessary. This kind of flexible, adaptive system offers the best of both worlds. It could become the new standard in enterprise software.
Make a move to build or adopt the Agentic AI revolution
We will soon have companies building specific Agentic AI solutions for other companies. We will also build our own Agentic AI.
Hence, the journey from SaaS to agentic AI is not just a technological upgrade.
All this is transform how we work and build or run a business.
AI agents are taking on the heavy lifting of everyday tasks. This allows human creativity and strategic thinking to have more room to flourish. The result will be organizations that are leaner, more agile, and more innovative than ever before.
As we look to the horizon, it becomes clear that the debate is not whether AI agents will replace SaaS. The question is how swiftly and smoothly this transition can be managed. The path ahead involves a blend of technological innovation, strategic foresight, and a commitment to continuous learning and adaptation.
Whether you are a business leader, an IT professional, an investor, or mere mortal human, the message is clear.
The age of agentic AI is upon us.
Those who prepare today will reap the benefits tomorrow.
How are you making the most of the AI Agents opportunity?
Do share in the comments how you are preparing or entering the AI Agents landscape.
At Applied AI Tools, we aim to make learning how to use AI software accessible. This includes many available AI tools for your personal and professional use. If you have any questions – email to content@merrative.com and we will cover them in our guides and blogs.
Learn more about AI Agents and AI tools:
- Learn about Stanford’s $50 trained S1 AI model – read
- What is OpenAI o3-mini? – read
- Best AI Meeting Assistant for project managers – read
- Best AI Meeting Assistants for sales – read
- Check out Prompt Engineering communities worth joining
- Explore ChatGPT alternatives for improved usability and productivity.
- Understand DeepSeek’s innovative changes that started the race to cost-effective AI engineering – read
You can subscribe to our newsletter to get notified when we publish new guides
This blog post is written using resources of Merrative. We are a publishing talent marketplace that helps you create publications and content libraries.
Get in touch if you would like to create a content library like ours. We specialize in the niche of Applied AI, Technology, Machine Learning, or Data Science.
Leave a Reply