--- title: LinearB Launches Agentic Studio to Help Engineering Leaders Measure Developer Productivity With AI description: LinearB's Agentic Studio answers engineering leaders' questions on AI coding. Heavy AI users merge 96% more code, while agent changes wait 5x longer for review. author: Darie Nani (Editor-in-Chief) date: 2026-11-02T14:16:17.457Z updated: 2026-11-02T14:18:23.203Z canonical: https://www.sovereignmagazine.com/article/measuring-developer-productivity-linearb image: https://cdn.nanimediahouse.com/linearb-agentic-studio-answer.webp categories: Artificial Intelligence, Startups content_type: Spotlight region: Boston publication: Sovereign Magazine about: - type: Organization name: LinearB description: LinearB is a developer productivity platform that turns data from the tools engineering teams already use into answers for engineering leaders and context for AI coding agents. Ori Keren and Dan Lines founded the company in 2018, and Gartner named it a Leader in its 2026 Magic Quadrant for Developer Productivity Insight Platforms. url: https://linearb.io foundingDate: 2018-01-01T00:00:00.000Z industry: Software sameAs: - https://www.linkedin.com/company/linearb --- Developer productivity platform LinearB today launched Agentic Studio, which lets engineering leaders ask plain-language questions about their teams' work, such as where delivery is stuck or whether AI is paying off, and get back an answer and a suggested next step. The coding agents that now write a growing share of companies' software receive the same information, including the past incidents, ownership and rework history of the code they are about to change. LinearB's research on 2.7 million pull requests, the proposed code changes developers submit for colleagues to review, found that the typical developer who used AI on at least three quarters of their coding days merged 96% more of them each week than a year earlier. The typical developer who did not use AI merged 3.6% fewer. ## AI Coding and Developer Productivity in LinearB's 2026 Benchmarks The AI figures come from 83,000 developers at 253 engineering organizations between February and May 2026. Even at the organizations LinearB ranks highest, AI helps write only 54% of pull requests. LinearB's larger 2026 Software Engineering Benchmarks Report, drawn from more than 8.1 million pull requests by 4,800 teams in 42 countries, found that AI-assisted pull requests were 2.6 times the size of unassisted ones at the 75th percentile. Pull requests opened by autonomous coding agents waited more than five times as long for review. At the organizations where agents are used most, they open 4.7% of pull requests. At the top-ranked organizations, 79% of agent pull requests are merged within 30 days, compared with 92% of those written by people alone. At organizations in LinearB's lower "fair" band, the figure for agents falls to 37%. LinearB says the likely reason is ownership, since a pull request from an agent that no developer takes responsibility for often never merges. Pull requests that receive an AI review merge at the highest rate of any type and raise the share of work that merges by up to five percentage points. At the 90th percentile, spending on tokens, the units AI providers charge by, reached $35.20 per developer per coding day, up nearly 75% in three months, while the median barely moved. ## Measuring the Return on AI Coding Tools Google's 2025 DORA report found that 90% of the roughly 5,000 technology professionals it surveyed use AI at work, up 14 percentage points on a year earlier. Boards and chief financial officers now ask engineering leaders whether the money spent on AI tools is paying off. "I use LinearB data and charts directly in my board decks, because my board is specifically asking questions about our use of AI and what our ROI is," said Ken Ringdahl, chief technology officer of expense management company Emburse. > "A dashboard says your cycle time is two days. So what? Now a leader can ask what to improve in their engineering organization and get a clear answer." > — Dan Lines, co-founder and COO of LinearB Leaders using Agentic Studio can build their own metrics, reports and dashboards with help from AI assistants in the platform, set which information each role can see, and pull the same data into outside AI assistants such as Anthropic's Claude through the Model Context Protocol (MCP), the standard many AI tools use to connect to outside data. LinearB says the AI assistant and MCP access have the highest user retention of any features it has released. LinearB builds those answers from data it collects from the tools engineering teams already use. It combines records from different code repositories and project management systems into one model, matches each engineer's accounts across those tools, and links every change from the original ticket to its release. It also records which work was done with AI and compares each figure with industry benchmarks, so leaders can tell whether a number is good or bad. Crypto exchange Kraken used LinearB's MCP server with an AI assistant to find out what was slowing its code reviews. The delays came from time zones and reviewers' availability rather than from the size of the changes. LinearB says Kraken more than halved review time at the 90th percentile within four months. "That's the structural shift," said Nik Sudan, engineering operations lead at Kraken. "It moves leaders from anecdote-driven to evidence-driven." ## Context for AI Coding Agents and LinearB's Effort Skill [Coding agents](https://www.sovereignmagazine.com/article/gartner-2026-magic-quadrant-enterprise-ai-coding-agents) start each task with no knowledge of a codebase's history. Through Agentic Studio they receive the same information leaders see, through MCP and through add-ons LinearB calls agent skills. "Everyone else is focused on writing code," said Ori Keren, chief executive and co-founder of LinearB. "We realized we hold a knowledge graph of everything happening across the software delivery lifecycle, and that's exactly what developers and agents need. So we rebuilt LinearB around it." The first skill, now in early access, sets how much effort an agent puts into each change. It rates a change's complexity, risk and importance to the business using the rework rate of the code involved, recent service incidents and who owns that part of the codebase. Simple, low-risk changes get less effort and use fewer tokens, while complex, risky or high-priority changes get more effort and careful, tested edits. The skill works with any coding agent, and leaders can see how often agents were routed to low, medium and high effort. Leaders can also set limits on what agents are allowed to do. LinearB's next skills will focus on reviewing and merging code, the stages where AI-written changes are now held up, and on helping companies govern AI-generated code and cut what they spend on AI. Agentic Studio is available to all LinearB customers from today. ## LinearB's Founders and Gartner's Magic Quadrant for Developer Productivity Ori Keren and Dan Lines founded LinearB in 2018. They met in 2012 at CloudLock, the cloud security company Cisco bought in 2016, where both held senior engineering roles. Lines also hosts Dev Interrupted, a podcast for engineering leaders that LinearB sponsors. The company raised a $50 million Series B in 2022, led by Tribe Capital with Salesforce Ventures, Battery Ventures and 83North, taking its total funding to about $71 million. In May, Gartner named LinearB a Leader in its first Magic Quadrant for Developer Productivity Insight Platforms. Jellyfish and Atlassian DX were also named Leaders. Atlassian agreed to buy DX for about $1 billion in September 2025. **About LinearB** LinearB is a developer productivity platform that turns data from the tools engineering teams already use into answers for engineering leaders and context for AI coding agents. Ori Keren and Dan Lines founded the company in 2018, and Gartner named it a Leader in its 2026 Magic Quadrant for Developer Productivity Insight Platforms. [Website](https://linearb.io) ## FAQ **Q: How do you measure developer productivity?** Most engineering teams track how many pull requests they merge, cycle time, how long reviews take and developer surveys. LinearB's 2026 data suggests that counting merged pull requests alone can mislead once AI is involved, because AI-assisted changes are larger and agent-written changes wait more than five times as long for review. LinearB links each change from ticket to release, records which work was done with AI and compares each figure with industry benchmarks, and Agentic Studio lets leaders ask questions about that data in plain language. **Q: What are some effective AI tools for developer productivity?** Teams use AI coding assistants and autonomous coding agents to write code, and AI review tools to check it. In LinearB's 2026 benchmarks, pull requests with an AI review merged at the highest rate of any type and raised the share of work that merges by up to five percentage points. LinearB's own effort skill, now in early access, works with any coding agent and sets how much effort and how many tokens the agent spends on each change according to its risk. **Q: What is the 40/20/40 rule in software engineering?** It is a rule of thumb that splits a project's effort into roughly 40% planning and design, 20% writing code and 40% testing. Most AI coding tools are aimed at the writing stage, and LinearB's figures show delays further along, with larger AI-assisted changes and agent pull requests that wait more than five times as long for review. LinearB's next agent skills focus on reviewing and merging code.