LinearB’s Agentic Studio tackles AI coding’s review bottleneck
More AI-written code is piling up in review queues faster than humans can clear it. That is the core problem LinearB just shipped Agentic Studio to address.
The Benchmark Numbers
LinearB’s 2026 Software Engineering Benchmarks Report covers 8.1 million pull requests from 4,800 teams across 42 countries. The productivity gains for heavy AI users are real: developers who used AI on at least 75% of their coding days merged a median 95.7% more pull requests per week in May 2026 compared to June 2025. Developers who used no AI at all merged 3.6% fewer over the same period.
The gains scale with usage. Developers using AI on half their coding days gained 65.4%. Occasional users gained 24.4%. LinearB notes these figures show a correlation, not proof that AI alone caused the gains, and acceptance rates for heavy users slipped 4.4% over the year.
The bottleneck is not output. It is what happens after the code is written.
- AI-assisted changes run to 408 lines of code at the 75th percentile, versus 157 lines for human-only changes. That is 2.6 times larger.
- Agent-written changes wait 17.6 hours for a first reviewer, versus 3.4 hours for human changes. That is 5.2 times longer.
- Across the full dataset, 84.4% of human-written changes were accepted within 30 days. Only 32.7% of AI-written changes hit that mark.
- At elite organizations, autonomous agents open just 4.7% of all changes, AI helps write 54% of pull requests, and AI writes 45% of the code that gets merged.
What Agentic Studio Does
LinearB already collects data from GitHub, Jira, and similar tools, linking each code change to the ticket that requested it and the release that shipped it. Agentic Studio exposes that history in two directions.
Engineering leaders can ask plain-language questions, such as where delivery is stuck or whether AI spending is paying off, and get an answer with a suggested next step. Coding agents get the same context through an MCP server, plus what LinearB calls agent skills. Before an agent edits a file, it can now see who owns that code, how often it has been rewritten, and whether it has caused recent incidents.
The first skill, called effort, is in early access. It uses delivery history to decide how hard an agent should try on each change. A low-risk edit gets a quick pass that uses fewer tokens. A complex or risky change gets careful, tested edits that use more. The skill works with any coding agent. The next skills will target code review and merging.
The Token Cost Problem
At the 90th percentile of spenders, developers now use $35.20 of AI tokens per coding day, up nearly 75% in three months. That works out to roughly $481 a month or $5,772 a year per developer, which LinearB says is still under 4% of total developer cost.
Elite organizations merge 2.0 times the pull requests of fair organizations, but spend $481 per developer each month on AI tokens against $50 at fair organizations. That is 9.6 times the spend for twice the output. LinearB’s worked example (an illustration, not customer data) shows a 350-developer company that quadruples token spend from $300,000 to $1.2 million a month while lifting output from 3.3 to 5.67 merged changes per developer each week. The result is a 29% drop in cost per delivered change, from $864 to $613.
Gartner expects spending on AI coding tools to exceed the average developer salary by 2028. The effort skill is LinearB’s attempt to slow that climb by routing expensive AI computation only to the changes that need it.
Early Customer Evidence
Kraken used LinearB’s MCP server with an AI assistant to diagnose slow code reviews. The finding was time zones and reviewer availability, not large pull requests. Over four months, Kraken cut its 90th-percentile review times by more than half.
Emburse CTO Ken Ringdahl puts LinearB charts directly into board decks. His board asks how the company uses AI and what return it gets. LinearB reports that its AI assistant and MCP access have the highest user retention of any features it has shipped.
Availability and Market Context
Agentic Studio is available now to all LinearB customers. Enterprises spent $4.0 billion on AI coding tools in 2025, up from $550 million a year earlier, making coding the largest single use of AI inside companies at 55% of the $7.3 billion spent on departmental AI, according to Menlo Ventures. MarketsandMarkets projects the AI code assistant market to grow from $8.14 billion in 2025 to $127.05 billion by 2032, a roughly 48% annual rise. Gartner sizes the developer productivity insight platform category at about $400 million, growing more than 40% a year.
LinearB CEO and co-founder Ori Keren describes the company’s data as a knowledge graph of the entire software delivery process that coding tool makers do not have. COO and co-founder Dan Lines says the goal is to move leaders from reading dashboards to knowing what to do next. 88.3% of organizations already use AI coding tools daily or weekly. 44.7% still do not measure what those tools deliver.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
