Reehan

Product @whatfix | Previously @browserstack

[Coffee with One 🇺🇸 @coffeewithone]

Let me tell you why Mintlify needs 50 people to "host .md files" and why 50 is actually too low.

I was the first intern at @mintlify. I sat three feet from Han and Hahnbee every single day. I watched this thing get built.

People see docs.stripe. com and think "oh, markdown renderer." That's like looking at Google and saying "oh, a text box."

Let me walk you through what's actually under the water.

You want search?
Not Cmd+K that returns garbage. Search that understands what a user means when they type "how do I authenticate." That's a whole retrieval pipeline. Embeddings. Ranking. Re-ranking. Edge caching so it feels instant in Tokyo and São Paulo. That alone is a team.

You want self-updating docs?
That means Mintlify is watching your codebase, detecting when your API changes, and flagging docs that are now lying to your users. Surprise! That's not a cron job anymore. That's diffing, parsing, mapping endpoints to prose, and doing it without false positives that destroy trust. That's another team.

WYSIWYG editing?
Sounds simple until you realize you're building a real-time collaborative editor that outputs clean MDX, not the garbage HTML that every rich text editor loves to produce. You're fighting ProseMirror. You're fighting the browser. You're fighting every edge case where someone pastes from Google Docs and injects 50 nested span tags.

Hahnbee taught me everything I know about engineering in those wall, and half of what she taught me was how to wrestle with exactly this kind of problem. The type safety was less about being academic and more about survival. One wrong type and the editor breaks for 10,000 companies.

Custom components?
That means shipping a component library wuth interactive API playgrounds, code blocks with syntax highlighting for 60+ languages, tabbed containers, callouts, cards.

BTW that has to render identically in the editor, in the build, in SSR, in the preview. Four rendering contexts. One source of truth.

If you've ever tried to make a React component behave the same in SSR and client-side, you know that's a PhD thesis disguised as a feature.

Authentication. Gated docs. Role-based access. SSO?
That means Mintlify is now in the auth business, which means they're in the security business, which means SOC 2, pen testing, token rotation, session management.

For docs.

AI analytics.
Not pageview counters. Understanding which docs are confusing users, which searches return nothing, where people rage-quit. That's event pipelines, ML models, and dashboards that have to make sense to a DevRel person who doesn't know what a funnel is.

SEO/GEO.

Mintlify doesn't just host your docs. They make your docs rank. Structured data. Sitemap generation. OpenGraph images generated on the fly. Meta tag optimization. Performance scores that stay green when you have 4,000 pages. That's infrastructure.

MCP servers. CLI tooling. Content checks that lint your docs like ESLint lints your code. CMS for non-technical writers to ship without a deploy.

And I'm not even going to get into the other hundred things. Versioning. Multi-language support. Custom domain provisioning with automatic SSL. Git sync that doesn't corrupt on merge conflicts. Preview deployments for every PR. Broken link detection across your entire site graph. Rate limiting on the API playground. WebSocket handling for real-time collaboration. OG image generation that actually respects your brand fonts. Middleware for custom routing logic. MDX compilation that doesn't choke on edge cases. Custom CSS injection without breaking the component tree. Cache invalidation, which, if you know, you know, across a globally distributed CDN.

Each one of those is a rabbit hole. Each one has a person at Mintlify who has lost sleep over it.

I watched founders of Mintlify obsess over this.

@handotdev would be the last person to leave at night and the first person in the office the next morning. He'd find a 200ms latency spike in the build pipeline and lose sleep over it. I watched him rewrite the entire settings page once. He did it not because it was broken, but because a user had to think for two seconds about where a toggle lived. He tore the whole thing apart and rebuilt it so that every section, every label, every grouping made immediate spatial sense. You open it, you know exactly where everything is. No customer filed a ticket for that. The culture of Mintlify is refusing to ship anything that makes a user feel lost, even for a moment, even on a page most people visit once.

@hahnbeelee was the same. Not only she taught me everything about Engineering I know today, she also taught me why things were built the way they were. Why this abstraction was chosen over that one. Why we don't take shortcuts even when the deadline is tomorrow. Every PR review was a lesson in caring about things that users would never consciously notice but would absolutely feel.

We moved fast. Extremely fast. But we cared.

A lot.

About things most people would never see.

The spacing between elements in the sidebar. The animation curve on the search modal. The way code blocks handle overflow on mobile. The fallback behavior when a component fails to render. They were less about building features and more about the difference between docs that feel like a product and docs that feel like an afterthought.

"But why can't you just vibe code it?"

You know who decided to use Mintlify instead of vibecoding?

@cursor_ai uses Mintlify.
@AnthropicAI uses Mintlify.
@Lovable used Mintlify
@twilio use Mintlify,
@perplexity_ai uses Mintlify
@Cloudflare use Mintlify

These are the most technical, most demanding companies on earth. They could build their own docs.

They have the engineers. They chose not to. Ask yourself why.

It's because docs infrastructure is a bottomless pit of complexity that has nothing to do with your core product. Every hour your engineers spend fixing a broken sidebar link or debugging why your OpenGraph images aren't generating is an hour they're not shipping features. Mintlify makes that whole problem disappear.

Vibe coding gets you a demo. It doesn't get you a system that serves 50 million page views without flinching. It doesn't get you an editor that 10,000 companies trust to not eat their content. It doesn't get you search that actually works. It doesn't get you infra that passes a SOC 2 audit. It doesn't get you the kind of reliability where Anthropic is comfortable pointing their entire developer ecosystem at your platform.

Mintlify is the infrastructure that looks invisible when it's working, which is exactly why people underestimate it.

"50 people to host .md files."

No. 50 people to build the platform that the best companies in the world trust with the first thing their developers see.

And honestly? 50 is actually too low.

[Zara Zhang @zarazhangrui]

Introducing "codebase to course", a skill that turns any codebase into an interactive coding course

So that you can learn coding through your own projects, complete with visualization, plain-English code translations, metaphors, even quizzes...

I vibe code a lot but have no idea how the code works under the hood. This is how I think "learning to code" should be in the AI age: Build first, learn later

Link below

[Deedy @deedydas]

Today, I'm excited to launch my lifelong passion project, Grand Old Books!! 🚀

There are 1000s of beautiful novels of the past, not in English, locked up in old PDFs, with no physical copies left. We started with Indian texts and brought back 12 books in 6 languages with pictures and annotations.

This is, and will always be, completely free.

We can't let time wash away history.

Please comment to let me know what book you'd like to see added.

[Reehan @reehan_6]

@idkohli “rol..touch grass”

[Raj Kunkolienkar @kunksed]

https://t.co/8XnigiADHc

[Ministry of Statistics & Programme Implementation @GoIStats]

NSO India unveils the MCP Server for eSankhyiki, enabling seamless integration of official statistics with AI tools.
Users can now connect directly to seven official datasets like PLFS, CPI, ASI, IIP, NAS & more through this beta version . Faster insights and smarter analysis through seamless access.
🔗 https://t.co/0fo6aXceYK
#AIReadyData #OpenGovernmentData #DigitalIndia #ViksitBharatBudget
@PMOIndia @Rao_InderjitS @_saurabhgarg @PIB_India @PibMospi @mygovindia @NITIAayog

[Reehan @reehan_6]

@kunksed so clawdbot is…the deep state?

[Nikunj Kothari @nikunj]

Alright, let me peel the onion on tech headlines that even smart journalists seem to be getting wrong on two areas: 1) revenue and 2) valuation (with note on ownership targets).

It’ll be a long post so grab a coffee..

Revenue
ARR (annual recurring revenue) as a metric is completely f*cked in 2025. Why? Historically ARR meant a subscription product that had a “seat” that a “person” would use with typically good retention and great margins.

The quality of this revenue was very good and you can pay high multiples (valuation divided by ARR) because you expected people to keep using it and more people to use it within an organization. This quality and multiples is important - keep in mind later.

In AI land, there are now a few types of revenue all lumped into “ARR”. Let’s break it down.

First, all now ARR is typically counted as annualized which means its calculating your last months revenue and multiplying it by 12.

Second, you’re seeing companies with NO subscription revenue report numbers as “ARR” (pick your favorite data labeling company but even random companies like a full body MRI). They call it annualized revenue run rate. This is last months revenue multiplied by 12. This revenue is NOT recurring and the quality is not great since margins typically are very very low (<20%). It’s a GMV/marketplace business similar to eBay where you get a cut of the revenue. Public multiples of these companies are far far lower than true ARR companies. So why do they call it ARR? Aha! It’s so they can say they’re the fastest to random revenue benchmarks and their valuation is compared to a software company vs lower multiples as they should. Growth rates are very high but the question is durability.

Third, you’re seeing a new class of AI companies that have a hybrid business model of subscription + usage (pick your favorite vibe coding platform). These companies are lumping all revenue as “ARR” including the usage component. So let’s say a company pays for 20 seats at $20/mo and collectively they spend an extra $200 on usage then calculating monthly revenue as $600 and multiplying by 12. Two reasons this is a problem: 1) The usage is NOT recurring and 2) Margins on those usage are typically low and will keep getting worse. Retention of this subscription is also not great fwiw. But, it makes the graph look great in the short term and allows them to raise on again SaaS multiples.

So now hopefully you understand now that all “ARR” is not equal - don’t even get me started at free trials and people fudging numbers by multiplying daily x 365.

If you’re a journalist or someone joining a company, here are smart questions to ask:

1) If it’s a regular subscription company, ask what % is annual and what is monthly? Historical retention especially on the monthly bucket. Expansion on key accounts.

2) If it’s a new type of AI company, then ask them to split their revenue and share exactly how it’s calculated. Ask them about retention in 12 month cohorts. And have them break out churn and why.

3) Spend time to learn about margin. Where are they actually spending money.

Every company I joined I spent time with the finance “person” and refused to join till I understood this.

Now, most companies won’t give you this since they don’t even give investors this but it’s your duty to ask and make sure your future equity is worth it.

Valuation in the upcoming post.

[Nikunj Kothari @nikunj]

Nuance is what makes us human..

[Varun Mayya @waitin4agi_]

Go watch!

[Karri Saarinen @karrisaarinen]

People talk about how AI is going to make design obsolete, and/or make pixel perfect designs not be a thing anymore. I don't think so.

Pixel-perfect design mostly existed in designers’ minds anyway. The mocks might perfect but the final product rarely was. Most of the time, those designs were shipped sloppily because the product organization didn’t have the patience to see the polish through. How many truly pixel-perfect products do you see out there?Especially with growth focused companies. There are exceptions, but not many.

The idea that AI might ruin visual quality feels like a non-issue since wasn’t much quality to ruin in the first place.

I also don't believe AI makes design obsolete but I believe it will raise the floor. That it’ll lift the skill level of product teams, and I hope it will free up time for the kind of polish that usually gets cut. So ideally we continue see more overall better design and the more "handcrafted" polished designs as well.

My general view of AI is that it will just let us do more things, not take away things. We didn't stop writing less when email was invented.

Few aspects:
1. AI as a skill multiplier. LLMs can elevate frontend and design quality for companies that historically couldn’t hire strong talent—or for individuals who haven’t built those muscles yet. In a way that way, it’s not that different from using UI kits, Tailwind, or shadcn.

2. Rethinking design systems. Design systems were a product of the ZIRP era, when teams scaled quickly and you couldn’t trust every person to design and build a decent button. So systems enforced quality through components and rules. But AI flip can this, and Instead of assembling rigid blocks, you can quickly build good scaffolding and refine with AI or by hand. LLMs might even enforce standards even better than design systems because they could be trained to spot inconsistencies and fix visual bugs automatically. The kinds of things that usually get deprioritized.

3. Designing closer to code.
I think AI will allow us to design more in code. I think it’s a good thing if we move away from pixel-perfect Figma files. The way I’ve always designed: Figma is where you design the vibe. Code is where you make it perfect. The real product is the one in production, not the mock. So that’s where polish should happen. Future design tools should make that process easier.

4. Taste still matters. For teams that already care about design, teams like @linear, none of this really changes the hard part. Achieving a polished UI is not that hard if you just have the practice. The hard part is conceptual. It’s figuring out how features fit together, how ideas map across the system. That’s where most of the iteration happens. That’s the part AI still can’t do for you.

So yes, AI will make things faster. It will increase the volume of output, but maybe it will also shift the baseline.

Holistic quality still depends on taste, systems thinking, and the willingness to care about the final experience.

[Reehan @reehan_6]

@kunksed What was your tinkering ratio? My limited but very recent experience with Replit Agent was sub par in terms of code accuracy - it didn’t even realise it needed to invoke API calls for the use case I described(via a detailed PRD, btw).

[Vishal Bhargava @VishalBhargava5]

Wow. Real Estate broker and prominent influencer @RafiqueMerchant says “We, Muslims must introspect why no one wants to give us a house”

Says “every community has some bad apples. Unfortunately in our community there are only some good apples”

[Trung Phan @TrungTPhan]

wow, found a rare interview of a DeepSeek co-founder talking about his first AI startup exit a few years ago

[claire vo 🖤 @clairevo]

This ended up worse than I thought.

Over 40% of people who responded work at places where less than 10% of the company uses their own product daily.

ngmi

My dogfooding non-negotiables:
- everyone has a production account, use it in “real life” scenarios
- leaders MUST actually use the product on a regular basis
- everyone on the team should be able to do a basic demo
- PM/EM/Design leads cannot source testing to engineers, they must also go end to end through new products personally
- have an opinion on the product, share it
- no one gets excused by “I’m not the target user”
- If you PM a dev tool, code is the product so you better learn to write code if you don’t already
- dogfood the whole service, including support, customer success, activation emails, docs, etc.
- come at everything with a unforgiving newbie eye, not your jaded expert eye
- if you’re paying for a competitor…yikes!
- but if you haven’t used your competitor’s product deeply (presuming you can)…yikes!

Too many teams treat their products like they’re theoretical, whatever is in a deck or on the website is enough to do the job.

But dogfooding is life. Everyone should be doing it. Leaders should be demonstrating it.

Why aren’t you?

[Reehan @reehan_6]

@rushabtated4 👋

[Gokul ⚡️ @gokulns]

(What I consider good) Career advice that no one asked me for: Over the past few weeks, I've been trying to crystallize some of my thoughts around winning in a job/career. Thought of putting this out there, in case it helps someone.

[Shreyas Doshi @shreyas]

Since time immemorial, when a CEO asks a PM at Product Review, “what do you need to 10X users/revenue?”, “what will make you go faster?”, etc

The PM steadfastly responds “We need [N] more engineers”. The Eng Mgr nods approvingly

A story thread, with some hard truths to swallow:

[Reehan @reehan_6]

@pranavmanie @navjha coincidence?? i think nottt

[Omang Agarwal @mango_sax]

@rapidobikeapp I was held hostage by your rider for hours in a dark patch outside my house.

I've a broken leg &amp; lung issue. He didn't let me drink water &amp; I obviously couldn't run.

I tried everything but had to ring the police. Your SOS doesn't work.

Shame on you