TL;DR
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Software engineer Neil Kakkar reports merging about 11 times as many pull requests per month at AI-native startup Tano as he did at PostHog before the rise of coding agents. His account describes faster output alongside new costs, including context switching, weaker shared understanding and less time for design; it is one engineer’s experience, not a controlled industry study.
Software engineer Neil Kakkar says he merged about 11 times as many pull requests per month at AI-native startup Tano as he did at PostHog before coding agents became part of his workflow. His account, published by The Pragmatic Engineer as the winning essay in a workplace-change contest, offers a firsthand example of how AI tools can raise an engineer’s shipping pace while creating new coordination and focus problems.
Kakkar compares his work at Tano, where he was a founding engineer, with four years at PostHog and earlier work at Bloomberg. He says he completed a similar breadth of product work at Tano in eight months that had taken 45 months at PostHog. At Tano, his projects included an affiliate platform, improvements to a content-monitoring system and a migration of the company’s agent architecture.
The reported pull-request figures are personal output measures, not a controlled productivity study. Kakkar says he merged 11 times as many PRs per month at Tano as at PostHog, and 2.4 times as many at PostHog as at Bloomberg. He cautions that PR counts can be gamed, while saying he did not set out to manipulate his numbers. The source does not provide the underlying monthly counts in the supplied material.
He attributes much of the difference between PostHog and Tano to AI tools and engineering culture. At PostHog through late 2024, he says, AI use was largely limited to autocomplete; at Tano, he joined a team already using coding agents. He also describes the earlier workplaces as requiring peer review and substantial review work, while his Tano workflow removes some communication and review overhead. The tradeoff, he writes, is that moving faster can mean breaking things and demands changes to how a team works.
Higher Output, New Workflow Costs
Kakkar’s account highlights a management question facing software teams adopting coding agents: how to capture faster delivery without losing coordination, design quality or shared context. If an engineer can produce more code and PRs, review systems and team communication may need to adapt rather than simply preserve processes built around slower individual output.
The account also shows why a higher PR count is not, on its own, proof of better productivity. Pull requests differ in size and impact, and Kakkar does not present a controlled comparison that isolates AI from other factors, including company size, role, experience or project scope. His figures are useful as a documented personal example, but they do not establish that other engineers or organizations will see similar results.
For workers, the reported gains come with demands on attention and stamina. Kakkar says faster work has brought more context switching, reduced shared context, less focus on design and concern about AI usage limits. Those costs matter because software teams must maintain systems and make sound product decisions, not only increase the volume of code shipped.
Popular right now
Hisense 65-inch U6 Pro Mini-LED 4K UHD Smart Fire TV
De'Longhi Dinamica Plus Fully Automatic Espresso Machine
Shark AI Ultra Robot Vacuum, Black/Silver, AV2511AE
As an affiliate, we earn on qualifying purchases.
From Autocomplete to Coding Agents
Kakkar’s comparison spans three workplaces with different sizes and cultures. He describes Bloomberg as a large financial technology company with about 20,000 employees and established processes. PostHog was a smaller, fast-moving startup when he joined, growing from around 10 people to about 200 during his time there. Tano had around 10 people when he wrote the essay, and he joined as its first engineer.
He treats his PostHog work as a pre-AI baseline, while acknowledging that PostHog’s use of AI has since advanced. He says that by late 2024 his AI tools there mainly provided code completion and helped with repetitive tasks such as writing test cases. At Tano, he describes working with coding agents capable of handling larger portions of feature development. The comparison therefore reflects different periods and workplace setups, not simply a before-and-after test of one tool.
The Pragmatic Engineer said it received 125 submissions to its essay challenge, announced around the publication’s fifth anniversary, and selected Kakkar’s piece for a $10,000 grand prize. The publication said it planned to feature runner-up essays in a later issue.
“I’m shipping 11x more PRs at Tano than at PostHog”
— Neil Kakkar, in his essay published by The Pragmatic Engineer
Limits of the Output Comparison
The essay does not establish how much of the reported change came from AI rather than differences in company size, job responsibilities, team practices, project mix or the engineer’s growing experience. It supplies no independent audit of the PR counts, and the source material provided here does not include the raw monthly totals or a definition of what counted as a completed unit of work.
It is also unclear whether the higher PR pace led to better outcomes for users or the business, or how much additional rework or maintenance it required. Kakkar presents the figures as evidence from his own experience, not as a result that can be generalized across software engineering teams.
Further Essays and Team Changes
The Pragmatic Engineer said it would publish or link to runner-up essays in a follow-up issue, offering additional perspectives on changes in technology workplaces. Kakkar’s essay does not set out a formal next test of his productivity claims or report a company-wide measurement at Tano.
For now, the next development described in the account is how teams adapt their working practices as AI agents become more capable. Kakkar argues that teams need broad participation and room to experiment to gain the benefits, while recognizing that existing tools and processes may not fit the new pace. Whether those changes can preserve code quality and shared understanding remains an open question.
Key Questions
What is the main development in the report?
The Pragmatic Engineer published software engineer Neil Kakkar’s essay comparing his work before and after joining AI-native startup Tano. He reports merging about 11 times as many pull requests per month at Tano as at PostHog.
Does the essay prove AI makes engineers 11 times more productive?
No. The figure is Kakkar’s personal comparison between two jobs with different teams, periods and responsibilities. The essay does not isolate AI’s effect or provide a controlled, industry-wide productivity study.
What costs does Kakkar associate with the faster pace?
He describes context switching, loss of shared context, less attention to design and concern about AI plan limits. He also says less communication and review overhead can mean teams move faster while risking more breakage.
What does Kakkar say changed between PostHog and Tano?
He says AI at PostHog through late 2024 was largely limited to autocomplete and repetitive tasks, while Tano used coding agents that could take on larger parts of feature development. He also points to differences in engineering culture and workflow.
Source: newsletter.pragmaticengineer.com
