Shopify’s AI Phase Transition: 2026 Usage Explosion, Unlimited Opus-4.6 Token Budget, Tangle, Tangent, SimGym — with Mikhail Parakhin, Shopify CTO

April 22, 2026

AI Summary

5 min read

Shopify’s CTO Mikhail Parakhin shared internal data showing that daily active AI tool usage among employees has approached 100%, with a clear phase transition in December 2025 when model quality crossed a threshold. “It’s hard not to do your job now without interacting deeply with at least one tool,” he said. The company funds unlimited tokens for every employee, with a floor of Opus 4.6 or equivalent quality. Parakhin also detailed three internal systems—Tangle, Tangent, and SimGym—that are reshaping how Shopify builds, experiments, and optimizes.

The token explosion and its management

The most striking chart shows token consumption growing exponentially, but with a skewed distribution: the top 10% of users consume far more than the rest. Parakhin admitted this feels “not ideal” because it suggests uneven adoption, but he defended Jensen Huang’s controversial claim that high token usage correlates with productivity. “He’s directionally correct,” Parakhin said, though he warned against the anti-pattern of running too many parallel agents that don’t communicate. The real metric to watch, he argued, is the ratio of tokens spent on code generation versus those spent on expensive, high-quality PR review models. “Good models write code with fewer bugs than the average human, but since they write so much more of it, more bugs make it into production.” The bottleneck has shifted

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

Never miss an episode of Latent Space: The AI Engineer Podcast

Get every new episode summarized in your inbox — free, ~5 minutes to read.

No spam. Unsubscribe anytime.

What you'll learn

1 (00:03) **Intro & Shopify’s AI Phase Transition** - Mikhail Parakhin, CTO of Shopify, joins to discuss the company's explosive AI adoption and internal tooling.
2 (02:47) **The December 2025 Inflection Point** - Parakhin presents internal data showing a phase transition in AI tool adoption around December 2025.
3 (07:54) **Token Budgets & The Anti-Pattern of Parallel Agents** - Parakhin defends Jensen Huang’s token budget comments and warns against inefficient agent usage.
4 (10:55) **The PR Review Bottleneck & CI/CD Crisis** - The volume of AI-generated code is overwhelming traditional code review and deployment pipelines.
5 (17:56) **Tangle: The Third-Generation ML Experimentation Platform** - An introduction to Shopify’s open-source tool for running and sharing ML experiments.
6 (24:17) **Content-Addressed Caching & Network Effects** - The key efficiency gain comes from automatic deduplication across the entire organization.
7 (26:14) **Tangent: The Auto-Research Loop** - An agent that automatically runs experiments to optimize any measurable goal.

+ Full timestamped outline available in the app

Guests on this episode

Mikhail Parakhin

Show Notes

Early bird discounts for the San Francisco World’s Fair, the biggest AIE gathering of the year, end today - prices will go up by ~$500 tonight so do please lock in ASAP!

From near-universal AI tool adoption inside Shopify to internal systems for ML experimentation, auto-research, customer simulation, and ultra-low-latency search, Mikhail Parakhin joins us for a deep dive into what it actually looks like when a 20-year-old, $200B software company goes all-in on AI. We cover why Shopify has become much more vocal about its internal stack, what changed after the December model-quality inflection, and why the real bottleneck in AI coding is no longer generation, but review, CI/CD, and deployment stability.

We also go inside Tangle, Tangent, SimGym, which are three major AI initiatives that Shopify is doing to make experimentation reproducible, optimization automatic, customer behavior simulatable, and search and catalog intelligence faster and cheaper at scale. Along the way, Mikhail explains UCP, Liquid AI, and why token budgets are directionally right but often measured badly, why AI-written code can still increase bugs in production, what makes Shopify’s customer simulation defensible, and what he learned from the Sydney era at Bing.

We discuss:

* Mikhail’s path from running a major Microsoft business unit spanning Windows, Edge, Bing, and ads to becoming CTO of Shopify

* Why Shopify is talking more publicly about AI now, and why staying at the frontier has become necessary for the company

* Shopify’s internal AI adoption curve, the December inflection, and why CLI-style tools are rising faster than traditional IDE-based tools

* Why Jensen Huang is directionally right on token budgets, but raw token count is still the wrong way to evaluate engineering output

* Why the real unlock is not more agents in parallel, but better critique loops, stronger models, and spending more on review than generation

* Why AI coding can still lead to more bugs in production even if models write cleaner code on average than humans

* Why Shopify built its own PR