Chris Strahl, Founder and CEO of Knapsack – Interview Series

Chris Strahl is the co-founder and CEO of Knapsack, where he focuses on reshaping how modern digital products are built by aligning design, engineering, and product teams around a shared system of truth. With a background rooted in design systems and front-end development, he is also widely known for hosting the Design Systems Podcast, where he explores how organizations scale design, improve collaboration, and modernize digital production.

Knapsack is an enterprise design system and digital production platform that acts as a living system of record, connecting design assets, code, content, and documentation in real time. The platform enables teams to build and govern reusable, production-ready components, manage design tokens, and maintain consistency across complex digital ecosystems. By structuring design and UI data in a way that is scalable and AI-ready, Knapsack helps large organizations accelerate delivery, reduce duplication, and ensure brand and product integrity across teams and channels.

Knapsack emerged after years spent building design systems for large enterprises at Basalt, where recurring friction between design files, engineering workflows, and shipped code became impossible to ignore. What was the moment when that pattern became clear enough to justify launching a dedicated platform?

We built countless design systems at Basalt, and the pattern was evident: design files, engineering workflows, and shipped code all existed in separate universes. The result was not a single dramatic failure but a thousand repeatable losses: mis-sized buttons, inconsistent behavior, and style drift across properties that cost teams months of rework. We knew it was a real problem when we saw that those issues could not be fixed with better sync plugs or nicer documentation. They required a single authoritative system of record for design, code, and brand rules. That realization made it clear that a dedicated platform was necessary.

Moving from agency and consulting work into building a product company revealed a deeper problem that existing design-system tools and workflow platforms weren’t addressing. What was the foundational gap that shaped Knapsack’s earliest architecture and direction?

When we shifted from agency work to building a product, the core missing piece became obvious. There was no dependable, machine-readable system that captured components, constraints, and the synergy between designers and engineers. Existing tools focused on files or isolated repositories, but not on a living representation of a product’s true state, including components, theming, usage rules, and compliance metadata. We built Knapsack around a canonical system of record that is component-first, versioned, instrumentable, and able to integrate with both design tools and codebases. That conclusion shaped our ingestion model and the linking layer, ultimately leading to the Intelligent Product Engine.

The “canvas era” is giving way to living, code-connected systems. How do you define this shift, and what changes for teams when product creation moves from static files to continuously updated systems?

The canvas era treated UX as static artifacts, usually files passed between teams. The new era is driven by continuously updated, executable systems that reflect real implementation. The change for teams is significant. Instead of debating which file or branch is the source of truth, they work from a shared system that exposes the current state of components, tokens, accessibility constraints, and production behavior. This reduces ambiguity, enables automated validation, and supports agentic workflows that generate usable UI based on real components rather than approximations.

Agent-generated UI often fails without a system of record that reflects real components, rules, and constraints. Why is this anchoring layer essential for AI to produce enterprise-ready interfaces?

AI can synthesize layouts and copy, but it needs an authoritative vocabulary to produce enterprise-ready interfaces. The anchoring layer, which contains concrete components, props, constraints, tokens, and usage rules, gives AI the boundaries it must respect. Without it, agents hallucinate styles, ignore accessibility requirements, or generate code that does not match what engineering teams actually ship. With a real component graph and ruleset, agents produce outputs that are implementable, compliant, and consistent with brand standards. This is the difference between a pretty mock-up and a deployable interface.

As the Intelligent Product Engine developed, what proved most difficult about unifying design assets, code, brand rules, compliance requirements, UX patterns, and performance data into one coherent system?

The challenge is not a single integration, but rather a series of them. It harmonizes intent and reality across various representations, including design tokens in Figma, component implementations in multiple repositories, brand guidelines in legal documents, telemetry from production systems, and compliance metadata. Each of these lives in different formats, with different owners, and on different update cycles. Turning these signals into one consistent model required strong ingestion pipelines, conflict resolution rules, and a clear model for provenance and ownership. Teams need to know what changed, who made the change, and why it was made. Building that trust layer was the hardest part.

With AI now capable of generating increasingly complete interfaces, how do you see the roles of designers and engineers evolving inside human–agent workflows?

Agents will handle repetitive tasks, such as scaffolding pages, proposing accessible variants, and generating localized content. Designers will focus on strategy, experience intent, edge-case UX, and defining the constraints that drive good outcomes. Engineers will focus less on typing out every pixel and more on component correctness, runtime contracts, observability, and performance. Humans become curators and validators. We define the rules, review outputs, and determine what quality looks like. The highest-value human skills will be systems thinking and judgment.

After the Series A, what became the highest-priority focus areas for accelerating product development and enterprise adoption?

The Series A allowed us to accelerate in three areas. First, onboarding and ingestion, which enable enterprises to create a system of record in days instead of months. Second, the Intelligent Product Engine, including model-aligned capabilities that ensure generated interfaces respect brand and rules. Third, enterprise controls, such as permissions, auditability, and compliance hooks, ensure leaders feel confident in adopting Knapsack across large organizations. These are the levers that drive real-scale adoption.

Enterprise teams often struggle to move from static workflows to dynamic, agent-ready systems. What are the biggest obstacles, and how does Knapsack help organizations adapt?

Enterprises struggle with fragmented systems, ownership silos, regulatory constraints, and the high cost of keeping everything up to date. We help by making ingestion fast and deterministic, by modeling provenance and ownership, and by providing governance features such as permissions and audit logs. These tools enable teams to validate trust in automated workflows.

As product creation becomes increasingly automated, what new capabilities do you believe teams must develop to stay effective in an environment where AI is generating more of the foundational work?

Teams must develop stronger system-thinking skills, specifically the ability to author constraints, policies, and component contracts that agents can use. They also need better monitoring and validation practices, including observability into agent decisions, rollout controls, and Q&A frameworks for generated UI. Governance literacy becomes essential, particularly the ability to express compliance, accessibility, and privacy requirements in a machine-readable format. The organizations that succeed will be the ones that can codify policy and quality into their systems.

Looking five years ahead, how do you expect AI-driven product creation to evolve, and what position do you want Knapsack to hold in that next stage of the industry?

In five years, product creation will resemble composing services against a live component graph, rather than passing static comps between teams. Agentic tools will generate production-ready surfaces using policies, performance budgets, and brand constraints. My goal is for Knapsack to be the canonical system of record that agents and apps rely on to understand a company’s true UI primitives and rules. This includes deep integration with models and CI/CD, strong governance for regulated enterprises, and speedy onboarding for new teams. Knapsack should be the trusted layer for brand, behavior, and safety as companies allow agents to operate more autonomously.

Thank you for the great interview, readers who wish to learn more about modern design systems and scalable digital production should visit Knapsack.