Custom Software + AI Agents: The Best Alternative to B2B SaaS
B2B SaaS forces you to compromise on features and workflows. Custom software combined with specialised AI agents is emerging as the best alternative, here's how the shift will play out.
I'm increasingly convinced that custom software, combined with specialised AI agents, will be the path to replacing B2B SaaS, at least to some extent. For two decades, buying software was the rational default because building it was expensive. That equation is changing fast. As LLMs become proficient at writing and maintaining code, the trade-off between 'buy and compromise' and 'build and fit' is tilting toward build. This post explains why B2B SaaS is a compromise at its core, how the 'bring your own AI' transition starts, and why specialised AI agents that deeply understand your codebase will change who builds and maintains business software.
Why buying B2B SaaS has always been a compromise
Traditionally, buying software is a compromise. You pick the option that's the closest fit, not necessarily one that solves all your problems. Then you bend your workflows and processes around the software instead of the other way around.
Over time, this software grows and becomes bloated, while you might only need 30% of the feature set. You pay for the whole platform, train your team on parts you never use, and work around missing pieces with spreadsheets and manual steps. It can even get to a point where the software slows down progress instead of accelerating it.
Building software was, or at least used to be, costly. When you compare the two, the compromise of buying software made more sense in most cases. That cost gap is exactly what's closing now.
The hidden double cost: your LLM bill plus your SaaS bill
Most companies today already pay for one or more LLM providers, and also pay an additional usage fee for third-party software that makes API calls to the same LLMs. You're effectively paying twice for the same intelligence: once directly, and once through your SaaS vendor's markup.
The first step of this transition will be 'bring your own AI,' where customers plug their own AI or LLMs into pre-built software. This breaks the vendor lock on the intelligence layer and exposes a simple truth: the value was never the wrapper, it's the workflow and the context.
Custom software + specialised AI agents: how the replacement works
Eventually, as LLMs improve and become increasingly proficient at coding, a combination of custom, pre-built software and a specialised AI agent that thoroughly understands the codebase can ship all the necessary features. The cost of building and evolving software collapses, and with it, the core argument for buying bloated SaaS.
Think of Claude Code or Cursor as a generic software developer you hire from Upwork. They're good, but they're not your developer. You have to explain the problems you're trying to solve, and even then, they don't have the company context, business goals, or your best practices.
We'll get to a place where we have a specialised developer, an AI agent, who is an expert in your codebase, your business requirements, and your company context. It knows why a decision was made two years ago, which edge cases matter, and how a new feature should fit the existing architecture.
How the builder's role changes
The role of builders will mostly shift towards maintaining these agents and deeply understanding customer problems to solve. Less time writing boilerplate, more time on discovery, scoping, and making sure the agent's output actually serves the business.
This shift will show up first in internal-facing, workflow-heavy systems where context matters more than polish, HR software, internal tools, ops dashboards, approval flows. These are exactly the systems where off-the-shelf SaaS fits worst and where company-specific context creates the most value.
The comeback of open-source B2B software
I also feel there will be a comeback of open-source B2B software: generic software with basic feature sets that companies can build on top of. Instead of renting a bloated platform, you start from a solid open foundation and extend it in your direction.
Agents specialised in understanding that codebase will maintain it with the help of humans. The open-source project provides the shared starting point, the agent provides the company-specific depth, and the human team provides judgment about which problems are worth solving.
What this means if you run a business today
You don't need to rip out your SaaS stack tomorrow. But it's worth auditing which tools you use at 30% capacity, where you're paying twice for LLM access, and which internal workflows are bent out of shape to fit a vendor's data model.
For new internal systems, the default question is flipping from 'which SaaS should we buy?' to 'should we build this with an agent-assisted team?'. The companies that start building that muscle now, custom software foundations plus specialised AI agents, will own their workflows instead of renting them.
