TL;DR
Get business pricing on networking and server gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Software engineer Neil Kakkar says he merged 11 times more pull requests per month at AI-native startup Tano than at PostHog, based on his own work history. His account describes faster shipping enabled by AI tools and a small-team culture, alongside new strains such as context switching and reduced shared understanding.
Software engineer Neil Kakkar says he merged 11 times more pull requests per month after joining AI-native startup Tano than he had at PostHog, in a first-person account published by The Pragmatic Engineer. His comparison offers a measure of how AI tools and workplace practices can change an individual engineer’s shipping pace, while also describing tradeoffs that the raw count does not capture.
Kakkar reports that he moved from four years at PostHog to eight months at Tano, where he joined as its first engineer. At PostHog, he says he merged an average of 2.4 times as many pull requests per month as he had at Bloomberg, where he started his career. He describes the Tano rate as 11 times the PostHog rate. These are figures from his own work history, not an independently audited industry study.
His work covered a broad range of products at both startups. At PostHog, Kakkar worked on areas including event ingestion, analytics, experimentation, surveys and feature flags. At Tano, he says he built an affiliate platform, improved a content monitoring system and helped migrate the company’s agent architecture. The report says his earlier PostHog figures reflect a pre-AI period; at that time, he mainly used AI for tasks such as generating repetitive test cases.
Kakkar credits AI agents and engineering culture for much of the pace change. He describes Tano as a small team designed to work with AI agents, while his earlier workplaces relied more on peer review and established processes. The source report says the shift reduced communication and review overhead, but brought new pressures: frequent context switching, difficulty maintaining shared understanding, less focus on design and concern about AI usage limits.
Faster Shipping, New Engineering Costs
Kakkar’s account illustrates why pull-request counts have become a visible, if incomplete, measure of AI’s impact on software work. If one engineer can produce and merge changes more quickly, a small company may be able to build or revise more of its product with a limited team. His reported work spans several substantial product areas, rather than only small code edits, although the comparison does not establish how much of the increase came from AI itself.
The account also points to a management challenge: removing review and communication steps may speed delivery but can make it harder for teams to maintain shared context and design quality. More code activity is not by itself proof of better software, greater customer value or higher productivity across an organization. Kakkar’s description of context switching and pacing suggests that faster output can move the bottleneck from writing code to coordinating work and deciding what should be built.
Products Worth Considering
Anker SOLIX C300 Portable Power Station, 288Wh
Bose SoundLink Flex Bluetooth Speaker (2nd Gen) – Portable Outdoor Speaker with Hi‑Fi Audio, Waterproof and Dustproof, USB‑C, Up to 12 Hours
Amazon Kindle Scribe (64GB) with Premium Pen, Metallic Jade
As an affiliate, we earn on qualifying purchases.
From Bloomberg to AI-Native Tano
The essay was selected as the winner of an essay challenge marking The Pragmatic Engineer’s fifth anniversary. The publication says it received 125 submissions and awarded Kakkar the grand prize of $10,000. It said a later issue would feature runner-up essays. This is an individual account published as a contest-winning essay, not a controlled comparison of engineering teams.
Kakkar’s three workplaces differed in size and operating style. He describes Bloomberg as a company of about 20,000 people with defined processes, PostHog as a startup that grew from roughly 10 people to about 200, and Tano as a company that grew from its first engineer to around 10 people. He characterizes PostHog’s pre-AI practices as fast-moving but still reliant on peer review. He also cautions that PostHog has since adopted AI more extensively, so his comparison is between different periods and workplaces, not between current versions of the two companies.
The technologies also changed between the periods he compares. Kakkar says AI at PostHog, until late 2024, was largely autocomplete and assistance with repetitive tasks. By the time he joined Tano, he says, Claude Code Desktop offered agent capabilities that could take on larger coding tasks. The report does not provide a detailed account of tool settings, review practices, or the exact method used to calculate monthly averages.
“I’m shipping 11x more PRs at Tano than at PostHog, where I shipped 2.4x as many as I did at Bloomberg.”
— Neil Kakkar, software engineer, in his essay published by The Pragmatic Engineer
What Pull-Request Counts Cannot Show
The report does not establish that AI alone caused the increase. Kakkar changed companies, roles, team size, tools and engineering processes, and he compares an earlier period at PostHog with a later period at Tano. The source provides no independent verification of the counts or consistent baseline for other engineers.
Pull-request totals can also vary with how work is divided into changes and how often code is merged. Kakkar acknowledges that the metric can be gamed, while saying he did not try to manipulate it. The account does not report defect rates, maintenance costs, customer outcomes, time spent on review or whether the faster pace persisted beyond his first eight months at Tano.
It is also unclear how widely the described tradeoffs apply. Kakkar argues that wider adoption of AI across a team helps capture its benefits, but the essay does not compare teams with different levels of adoption or provide measured productivity results beyond his personal output.
More Essays and Broader Evidence
The Pragmatic Engineer said it planned to publish a follow-up highlighting runner-up essays and linking to other submissions. Those accounts may add perspectives on how work practices are changing, but they would still be individual reports unless supported by broader comparable data.
For Kakkar’s comparison, the next useful evidence would be whether his reported pace continues and how it relates to quality, customer impact and team coordination. The source report does not announce a formal follow-up measurement or provide a date for one. For now, its clearest finding is limited to Kakkar’s account: his recorded pull-request pace rose sharply after a move to an AI-native startup, while the reasons and overall effect remain difficult to isolate.
Key Questions
What change did Neil Kakkar report?
Kakkar says he merged 11 times more pull requests per month at Tano than at PostHog. The figure is his personal comparison, not an industry-wide estimate.
Does the account prove AI caused the increase?
No. Kakkar credits AI tools as a major factor, but he also changed companies, roles, team settings and engineering processes. The report does not isolate AI’s effect.
What tradeoffs did he describe?
He reports more context switching, difficulty maintaining shared understanding, less focus on design, harder pacing and concern about AI plan limits.
Does more pull-request activity mean better software?
Not by itself. The essay does not provide comparative defect rates, maintenance costs or customer results, so the output count cannot establish software quality or business impact.
Source: newsletter.pragmaticengineer.com
