Skip to content

A Post SaaS world notes

At Causify we are building tools for the post SaaS world, enabled by coding writing AI agents.

Trend 1: "build-vs-buy has flipped in many categories"#

Code-Writing AI Agents reached production-grade quality. For the first time, companies could feasibly ask: "Why are we paying Salesforce $500,000 annually when we can replicate 80% of the functionality with Claude Code in a day?"

old SaaS playbook — the one that relied on simply building a decent product, slapping on a subscription price, and expecting a steady stream of customers without ongoing strategic innovation

People buy SaaS because building custom software was prohibitively expensive

In https://www.anthropic.com/engineering/building-c-compiler I tasked 16 agents with writing a Rust-based C compiler, from scratch, capable of compiling the Linux kernel. Over nearly 2,000 Claude Code sessions and $20,000 in API costs, the agent team produced a 100,000-line

Yes — horizontal SaaS is being repriced.

Trend 2: Complete customization#

Deliver outcomes so complex that building it in-house would be prohibitively difficult (e.g., fraud detection, regulatory compliance)

Go deep in verticals incumbents ignore

Yes, that article caught our eyes since it clarified some our positions that we have been writing about almost 6 months ago https://blog.causify.ai/cracking-the-long-tail-of-data-science-problems.html https://blog.causify.ai/quote-of-the-day-ai-has-broken-wrights-law.html https://blog.causify.ai/your-data-isnt-as-ready-as-your-slide-deck-says.html

The problem is that VCs have been telling us "a platform is too horizontal, we want vertical applications", "too consultant-ey". All points that are valid in the old world, but are irrelevant (or wrong) in the post-AI-everywhere.

Outcome-driven economics — not feature subscriptions

Trend 3:#

AI-native from inception — not retrofitted We have developed our company around a "completely remote", "everything is documented", "everything is modular", "everything is unit tested" which is perfect for coding agents

We do see the 10x improvement in throughput from who embraces coding AI with strong engineering. At the same time the vibe coding approach is appealing ("a simple secret for ...", "a pill to solve your problem ...")

AI-first world

The market is rewarding capital efficiency and AI-first architectures.

Can we use AI agents to scale customer support, sales, or onboarding? Can we build features 10x faster with Claude Code or Cursor?

Capital-efficient scaling — not growth at all costs

What is our response?#

Causify is not horizontal SaaS. “We are building the decision engine layer for enterprises operating under uncertainty — not another SaaS workflow tool.”

Regime-sensitive, high-stakes, probabilistic decision problems Problems where wrong decisions carry real financial cost Problems AI agents alone cannot reliably solve

When we pitch, investors often react with skepticism: “Wait, you’re building tools for hedge funds, predicting wind turbine failures, forecasting just-in-time inventory for schools, and optimizing curtailment for power generation? You’re in too many markets. Focus on a thin slice that can scale a million times.” They’re right in one sense: venture capital thrives on focus. But history shows investors rarely predict which companies will become transformative. We’ll prove this in a future post when we apply Causify to the VC industry itself. For now, let’s take their comment at face value and debate it.

Here’s what’s missed: Causify is not in multiple markets. We are in one market: the market of causal platforms. All the problems listed share the same architecture and the same causal components, merely rearranged for different domains. Each one of those “niche” use cases is actually a hundred-million to multi-billion-dollar opportunity, depending on how the future unfolds. And while the conventional VC playbook would encourage a different startup for each domain, those startups usually overfit to their vertical, building siloed solutions that cannot transfer learnings. At Causify, we avoid this trap by learning cross-sectionally.

Consider the parallels:

  • Hedge funds predict alpha and risk using their understanding of the markets, using price histories, company disclosures, and macro data
  • Wind turbine operators predict component failures based on how turbines are built and operate, using sensor data and physics
  • Supply chain companies forecast Chromebook part needs using seasonality and past breakage rates
  • Power utilities forecast curtailment using how power markets work, based on market prices, demand, and energy grid data

Did you notice the common structure? Each problem reduces to the same form

  • Predict Y based on an understanding of how system Z works, using data X, and then make an optimal decision based on Y

The unifying insight is that all these problems can be decomposed into the same basic components:

  • Prediction of outcomes (machine learning to predict Y on features X)
  • Understanding of system dynamics (causal AI)
  • Decision-making under economic criteria (causal optimization)

Our moat is:

Causal tooling Decision optimization layer Domain calibration Complexity that is non-trivial to replicate Clear Messaging for Management

References#

Why SaaS Is Dead? And What’s Replacing It? — https://blog.saasvolt.com/why-saas-is-dead-and-whats-replacing-it/ — Apr 23, 2025  The AI Slow Roll Is Killing Your SaaS — https://www.saastr.com/the-ai-slow-roll-is-killing-your-saas-why-existential-dread-is-needed-today-the-latest-20vc-with-rory-harry-jason/ — Jun 2025 The AI Revolution in SaaS Is Here—But Won’t Arrive All at Once — https://www.newsweek.com/the-ai-revolution-in-saas-is-here-but-wont-arrive-all-at-once-11157980 — Dec 4, 2025  SaaS Isn’t Dead, It’s Just Having an Agentic Makeover — https://www.forbes.com/councils/forbestechcouncil/2026/01/16/saas-isnt-dead-its-just-having-an-agentic-makeover/ — Jan 16, 2026  $300 Billion Evaporated. The SaaS-pocalypse Has Begun. — https://www.forbes.com/sites/donmuir/2026/02/04/300-billion-evaporated-the-saaspocalypse-has-begun/ — Feb 4, 2026  The SaaSpocalypse of 2026: How Agentic AI Killed Per-Seat SaaS — https://www.outlookindia.com/xhub/blockchain-insights/the-saaspocalypse-of-2026-how-agentic-ai-killed-per-seat-saas — Feb 2026  $1T SaaS Selloff: Why AI Is Replacing Software — https://serenitiesai.com/articles/1-trillion-saas-selloff-ai-replacing-software-2026 — Feb 2026 The Sassocalypse: Why 1 Trillion $s in SaaS Market Cap Vanished in Feb 2026, "https://www.linkedin.com/pulse/sassocalypse-why-1-trillion-saas-market-cap-vanished-feb-madhuvarsu-mu3mc/", February, 2026

// https://docs.google.com/document/d/1ZQWmHAE48PNUy2KQltKX0FD4Yi_MQxgE4iuivD2upq8/edit?tab=t.wp8yw2s1c22a