If SaaS Is Dead, Linear Didn't Get the Memo
AI & I
Apr 01
If SaaS Is Dead, Linear Didn't Get the Memo
If SaaS Is Dead, Linear Didn't Get the Memo

AI & I
Apr 01
In this episode, Dan Shipper speaks with Karri Saarinen, cofounder and CEO of Linear, about how the company has thoughtfully integrated AI not as a bolt-on feature but as a foundational layer that augments human judgment and workflow coherence.
Linear’s AI strategy centers on deep workflow understanding—not chatbot gimmicks—leading to a purpose-built agent platform that guides, structures, and enhances software development. Rather than chasing AI hype, Linear spent two years refining its approach, prioritizing automation that fits naturally into engineers’ existing tools and processes. The company rejects the 'SaaS is dead' narrative, arguing instead that disciplined execution and first-principles thinking give agile startups an edge over rigid public companies. Linear measures AI impact through quality signals—like bug rates and token efficiency—not superficial metrics like PR volume. It emphasizes conceptual work before coding, using low-risk exploration to frame problems and validate AI tools rigorously. Its product strategy now embeds agents directly into PM and engineering workflows, enabling tighter context preservation and usage-based billing for advanced capabilities. A live demo shows how shared context, collaborative chat, and human-verified code reviews compress cycles without sacrificing quality. Ultimately, Linear affirms that while AI will grow more autonomous, human intent, judgment, and craft remain irreplaceable in building meaningful products.
00:00
00:00
Linear serves as a system for guiding agents and building context in the SaaS era
04:33
04:33
Simply integrating a ChatGPT-like chatbot isn't useful
05:10
05:10
Linear aims to understand workflows and what's valuable, adding a chat interface
07:46
07:46
We had no pressure from investors to launch AI features earlier—we waited for real value
15:03
15:03
Coding agents can do the first pass on bug fixes, and there's a good workflow in Linear for code review
20:37
20:37
Shortening the loop to make actions immediate rather than waiting
27:37
27:37
With AI, it's hard to determine the usefulness of tools due to non-deterministic LLMs
35:35
35:35
Usage-based billing is necessary for coding agents due to high token costs
46:16
46:16
Linear avoids being a 'kitchen sink' product and instead focuses on accelerating the natural next steps in the development workflow.
47:48
47:48
In the next five years, product development will have more self-driving aspects, with features acting like agents making decisions based on input and context