"SaaS is dying" — the AI-native custom software revolution
Why Palantir and Anthropic insiders are calling the end of the SaaS era, and why it's the biggest opportunity developers have ever had.
Every decade, the technology industry goes through one large, structural shift. In the 2000s, software came in shrink-wrapped boxes and a perpetual licence. In the 2010s, the cloud arrived and SaaS — Software as a Service — quietly conquered the world. We all got used to paying a monthly fee for a CRM, an ERP, a marketing suite we never actually owned.
Now a number of serious voices inside companies like Palantir and Anthropic are saying the quiet part out loud: the traditional SaaS era is winding down.
So why is an industry that generates billions of dollars a month in recurring revenue starting to crack? And more importantly: if off-the-shelf software dies, do developers go hungry — or are we about to enter the biggest gold rush for custom work the industry has ever seen? Let’s look at the shift through the lens of the global market.
Why off-the-shelf SaaS quietly exhausted businesses
Until very recently, every mid-sized and large business followed the same playbook. Need a CRM? Subscribe to Salesforce or HubSpot. Need to manage your supply chain? Buy a packaged ERP.
The model came with three quiet but enormous problems:
- “One size fits all” is a polite fiction. A packaged product never matches your business 100%. You end up reshaping your business to fit the software’s opinions instead of the other way around.
- Customisation costs are unbounded. Adapting a packaged system to a specific company routinely meant months of work and millions of dollars paid to an army of integration consultants.
- SaaS bloat is a tax. You pay for the 80% of features bundled into the suite that you’ll never touch.
The pattern that exhausted buyers wasn’t the price. It was buying software to solve a problem, and then having to spend more money making the software actually fit the problem. They want the answer, hyper-specific to their situation, the first time.
The AI-native shift: from “buy” to “build”
In the old world, writing software from scratch was expensive, risky, and slow. That’s the whole reason SaaS won — not because packaged was better, but because custom was worse. Today, that equation has flipped 180 degrees.
Claude Code, OpenAI Codex, and the new generation of agentic coding platforms have driven the cost of producing software down to nearly zero.
Old formula: off-the-shelf SaaS + months of integration = expensive, constrained
New formula: AI + one specialist developer + a few days = hyper-specific, cheap, ownedWhat used to be a quarter-long project — say, a custom internal supply-chain tool or a bespoke logistics dispatch — is now genuinely something one developer with the right tools can build from zero in a handful of days. The result matches the business to the millimetre, has no recurring subscription, and ends up as the company’s own intellectual property.
Where the new contracts are going to come from
The standard fear — “if AI writes all the software, why would anyone hire a developer?” — has it almost exactly backwards. The death of off-the-shelf SaaS opens the largest custom-software market in history. Three veins of work are already showing up.
1. “SaaS killer” projects
Small and mid-sized companies are no longer willing to pay $50–$100 per seat per month for Salesforce, Jira or Asana. They are starting to go to local and global freelancers and say: “Build me an internal tool, tailored exactly to how we work, with AI baked in — from scratch.”
2. Wiring AI agents into the business
Companies don’t just need someone who can write code — they need someone who understands the business and can hook AI agents into it. The most valuable contracts are turning into “AI architect” work: building internal systems that analyse the company’s own data and make decisions in real time.
3. Data privacy and sovereignty
GDPR and its global imitators have made companies genuinely nervous about handing customer data to a third-party SaaS. There is real, growing money in building secure, closed systems that run on the customer’s own infrastructure — on-premise by design.
| The old developer | The AI-native developer |
|---|---|
| Spends months chasing syntax and bugs | Stands up an architecture in hours with AI |
| Glues together other people’s APIs and SaaS | Builds hyper-specific solutions from scratch |
| Knows how to write code | Looks at a business problem and proposes the algorithm |
Who wins the global market?
Geographic borders matter much less in this shift. Large companies in the US and Europe are increasingly skipping local agencies — which charge millions — and hiring global developers who are genuinely fluent with Claude Code and the new tooling.
If you’re a developer, the strategy to win in this market is to change what you sell:
- Stop saying “I write code.” Writing code is now a commodity. AI does it too.
- Start saying “I solve the problem.” Talk to business owners in their language. They don’t care which framework you picked. They care how quickly and how cheaply you can deliver something that works.
- Get genuinely fluent with the tools. One developer with AI can now do the work of five. That means your speed — and your billing — can scale the same way.
Closing
SaaS is not going to vanish. But it is losing its monopoly. The future belongs to flexible, cheap, AI-native custom software built from zero for exactly one company.
The deeper shift is a democratisation of software. The dependency on a handful of packaged giants weakens, and independent developers and small innovative businesses become the new leaders of the digital economy. The rules of the game just changed — and as usual, only the ones who adapt first are going to win.