In this article
As AI rapidly evolves, it's becoming a core requirement for modern SaaS platforms. If you're building or scaling a software product in 2025, understanding how AI shapes user expectations, product development, and operational efficiency is critical.
In this post, we break down the most actionable insights from a recent webinar with our guest, Yariv Adan, Ex Senior Director at Google AI, Founding General Partner at ellipsis, and Leo Goldfarb, Managing Partner at Albato.
You’ll learn about the real-world impact of AI in SaaS, including the rise of AI agents, the role of the Model Context Protocol (MCP), and how teams can structure themselves to win with AI.
5 ways AI is changing the SaaS landscape
These are some of the upcoming trends that are making an impact in the SaaS industry.
1. AI capabilities are now a core user expectation
If you haven’t already, it’s time to start thinking about how AI can fit into your SaaS product—because your users are starting to expect it. AI features help people save time and reduce manual work. And once they get used to that kind of experience, it’s tough to go back to doing things the old way.
Chances are, many of your users have already seen AI in action—whether it’s natural language search, content generation, or smart recommendations. If your product doesn’t offer those kinds of time-saving tools, it might start to feel a bit behind the curve. And in a competitive market, that can mean more friction, higher churn, or even lost sales.
What you can do: Identify which of your product's activities and workflows are the most tedious and time-consuming—yet critical—for your users. These friction points are your best opportunities to introduce high-impact AI features that make a real difference.
2. AI agents are driving real business value
Yariv Adan’s venture fund, Ellipsis, is a great example of how AI agents can streamline and enhance complex core business processes—enabling a small team of VC partners to operate with impressive efficiency.
Rather than hiring a team of investment analysts, the VC fund strategically chose to leverage AI for automation. The AI-first approach works across the entire VC cycle—from sourcing new investment opportunities to conducting detailed due diligence for each one.
When a pitch deck is received, AI is used to extract and summarize the key insights quickly. From there, AI-driven analysis helps pre-vet pitches and cherry-pick the most promising ones.
Then, a crew of specialized AI agents—each focused on areas like team, product, market, and risk—contribute their findings. A “manager” agent brings everything together and delivers a complete investment committee scorecard with a clear recommendation—all in a matter of minutes.
Check out our post about AI agents if you want to learn more about how they work and how to use them in your business.
3. Some AI features might fail. Good news: you can avoid it
Many people think AI success is defined solely by how well the technology works. That’s a misconception. In reality, success is all about AI being relevant to your business goals and customer pain points.
Yariv has highlighted several common pitfalls he often sees when companies try to implement AI:
-
No clear business value. One of the biggest mistakes is building AI solutions without a well-defined problem to solve—or without any impact on core business metrics like revenue or cost savings. If it doesn't move the needle, it’s probably not worth doing.
-
Following the hype. Jumping into AI just because “everyone else is doing it” rarely ends well. Without a clear strategy, defined goals, or KPIs to measure success, it’s easy to waste time and money on tools you don’t need.
-
Poor quality or UX (for customer-facing agents). If AI agents are complex to set up or prone to hallucinations, people won’t use them. As obvious as it may seem, agents need to be user-friendly and predictable to achieve real adoption.
-
Overspending with little return. As Yariv puts it, using AI where it's not needed can be like “using a Lamborghini to deliver pizzas.” If the solution is over-engineered for the problem, the ROI will likely be disappointing.
-
Automating chaos. If your existing processes and data are unclear, inconsistent, or undocumented, attempting to automate them with AI will only exacerbate the issue. As the saying goes: garbage in, garbage out. Before you bring AI, make sure the processes and data are solid.
What you can do: Focus on measurable outcomes from the start. Define clear success metrics and ensure the processes you’re automating are worth the investment. That’s the key to making AI deliver tangible business results.
Guide by Albato Embedded
Top 10 Mistakes B2B SaaS Startups Make — And How to Avoid Them
Download
4. MCP is becoming a game-changer for AI integration in SaaS
MCP solves a key challenge: it provides a common standard for how AI agents interact with software tools.
While largely invisible to the end-user, MCP is a significant enabler for developers and agent builders, allowing them to build more interconnected AI applications.
For more details about Model Context Protocol (MCP) and why it matters, check out our blog post.
5. AI vs traditional automation: When to use what
AI agents shine when it comes to unstructured, open-ended tasks—like extracting data from PDFs, websites, or even screenshots. However, when dealing with predictable, repetitive workflows (such as sending an email after a form submission), traditional no-code automation remains the better fit.
The smart approach? Use both—each where it’s strongest.
Agents are great at handling messy inputs. They’re especially useful for tasks that require natural language understanding or involve navigating multiple possible paths to reach a solution. For example, Yariv highlights how agents can extract company names from PDFs, websites, or images and then organize that information into a structured format like JSON.
Once the data is structured, that’s where traditional automation takes over. Tools like no-code workflows are ideal for clearly defined, step-by-step processes—updating databases, triggering notifications, and so on. They’re reliable, easy to debug, and more cost-effective for routine tasks that don’t require complex reasoning.
In short: let agents handle the fuzzy part, and use workflow automation to add precision and consistency.
For more insights about how AI is changing the future of integrations, check out the full video on our YouTube channel.
How to build an AI-ready SaaS team
To succeed in the new era, companies must build their “AI muscle.”
What that means:
-
Create dedicated roles (e.g., Head of AI Transformation).
-
Build internal expertise across teams.
-
Invest in MLOps practices to scale, monitor, and optimize AI in production.
Businesses should consider combining AI agents, workflow automation, and human oversight (Human-in-the-Loop) to enhance their operations. The best combo depends on how much risk and precision a specific task needs.
Check out our article about risk-precision framework to choose the right approach for your task.
Summing up
AI is no longer a nice-to-have in SaaS. Whether you're enhancing UX, speeding up MVP development, or building automated workflows, the companies that embrace AI today are building the platforms users will switch to tomorrow.
Want to future-proof your SaaS product? Start by identifying where AI can deliver tangible value—and take one step forward today.
Albato can help take your SaaS product to the next level by making it easy to integrate powerful automation and AI workflows—with no code. With Albato, you can connect your app to hundreds of other tools, streamline manual processes, and enhance user experience through seamless data flow. Whether you're looking to automate onboarding, sync customer data, or trigger AI-powered actions, Albato gives you the flexibility to build smart, scalable systems that grow with your product.
Read more:
