Maveen Mushtaq

AI Agents for Public Health Insurances and Financial Services - Founder @9one3ai

[Maveen Mushtaq @maveenbhat]

User interfaces with good taste are gonna be a moat even in the AGI era.

[Maveen Mushtaq @maveenbhat]

I just started using @conductor_build. I really love it, was looking for something like this simple but so powerful.

My two asks to make it complete:
• Native notes tab, I always context-switch to track next todos
• Embedded @warpdotdev one less window to juggle

Just these two would make it my permanent dev workspace.

[Maveen Mushtaq @maveenbhat]

The next $1 trillion company will be a software company masquerading as a services firm.

The old question was: can we build this?
The new question is: why would we wait?

Few years ago, building something custom for a customer meant a conversation that started with “we’d love to, but…” months of dev time, architecture debates, prioritization tradeoffs.
That friction wasn’t a product decision. It was a technology constraint masquerading as one.

The constraint is gone.

A complex, custom workflow that used to take three months now takes a day. In some cases, an hour. Not because engineers got faster, because the nature of building changed. AI doesn’t just accelerate code. It collapses the distance between problem and solution.
What this means for services is underrated.

@JulienBek at Sequoia just wrote about it: the next trillion-dollar companies won’t sell software. They’ll sell the work. Not a tool that helps you close the books, a service that closes the books.
I think there’s something even more interesting underneath that thesis.

When customization was expensive, you had to standardize. You built one product for many customers. You made them adapt to your workflow. You questioned every edge case: is this worth building? Can we support this at scale?
That’s over.

Now you can build for the customer, their process, their language, their edge cases, and deliver it fast enough that “custom” and “scalable” are no longer in tension. Custom is the moat. Not custom that breaks when you update it, but custom built on a shared infrastructure that you own, that compounds with every client.

This is what engineers need to understand. Software development isn’t dying. But what we’re protecting has shifted. The value is no longer in writing the code. It’s in knowing what to build, ensuring it’s secure, designing the architecture that holds everything together as it scales.

Judgment, taste, and trust are the new scarcity.

The next wave of great companies will look strange at first. They’ll seem like service firms. They’ll price like service firms. But underneath, they’ll be software companies, and their moat will be the proprietary infrastructure nobody else has, plus a thousand bespoke solutions built on top of it.

[Maveen Mushtaq @maveenbhat]

@benln Interested

[Maveen Mushtaq @maveenbhat]

@benln https://t.co/0cWEAUrUqs

[913.ai @9one3ai]

🚀 Big news: https://t.co/ULqKQoMZgj’s domain-specific AI agents now run on GPT-5 on default!

Want early access?

You can use it for free here: https://t.co/uEb7wcaszs

Let’s redefine what your team can do when AI isn’t just assisting - it’s working.

[Maveen Mushtaq @maveenbhat]

Good morning and happy Sunday! ☀️

I’ve been thinking a lot about the current state of AI tools. We all know the story: There are plenty of platforms out there, whether it’s agents in @ChatGPTapp tools like @lovable, @cluely, @cursor_ai etc that just aren’t perfect yet. Sometimes they break, sometimes they miss the mark, and nobody would trust them (yet) with their most mission-critical, production workloads.

But here’s what I find fascinating: These companies are still winning, because they’re focused on distribution.
They’re shipping products that are “good enough” for prototyping and low-risk use cases. They’re getting feedback, building a user base, learning what works, and most importantly: they’re capturing the market.

Now, imagine what happens in a year or two, when LLMs like GPT-5, GPT-6, or whatever comes next make another quantum leap. The companies who’ve spent this “imperfect” era obsessing over distribution, user needs, and real-world problems will be sitting on gold. They’ll have the distribution, the user data, the brand recognition, and the operational knowledge to simply swap in a better model, and suddenly, their “not quite ready for prime time” tools become the platforms everyone must use for production.

Meanwhile, the folks waiting for the models to be “perfect” before launching anything will be left scrambling, trying to catch up.

It’s not about being flawless today.
It’s about being there, learning, and building for the moment when the tech finally catches up with the vision.

The future belongs to the companies building distribution now, not to those waiting for perfection.

If you’re building or thinking about launching, don’t wait for perfect. Get it out, get it in people’s hands, talk to users and that’s the culture we are trying to build @9one3ai as well.

[Maveen Mushtaq @maveenbhat]

Over the weekend I ran an experiment: I handed @cursor_ai agent a feature list, and 48 hours of autonomy. I kept about 20 % of what it wrote and deleted the rest.

I wrote up the details, numbers, mistakes, and what “trust” actually costs, in an essay: https://t.co/cdU13u9kFr

[Maveen Mushtaq @maveenbhat]

Just a year ago it was super difficult for LLMs to get the frontend code right, but now you can generate beautiful, functional frontends within minutes using tools like v0, replit etc and LLMs are improving incredibly fast and it’ll be exciting to see what’ll be possible in a few months (not years)

[Maveen Mushtaq @maveenbhat]

AI Won’t Replace You—But Someone Using It Might.

Today, I’d like to introduce the newest member of our https://t.co/ZsxiRdohUz team: Ava.

Ava isn’t just any team member—she’s a Digital Worker. She analyzes contracts, spots risks, compares agreements, and generates legal insights & reports in seconds. She works like a legal expert, always sharp, always improving, and available 24/7.

Ava doesn't just follow instructions; she learns, adapts, and iterates with you. Ava works with you—not just for you.

Ava is just the beginning.

If you want to see Ava in action or explore how Digital Workers can support your team, reach out. We’d love to show you what’s possible.

📩 DM me to try Ava or discuss your use case.

[913.ai @9one3ai]

🚀 AI Agents Are Changing Document Processing – Here’s How

Watch the full video here: https://t.co/wWzumBHH54!

Most businesses still handle documents manually—reading, sorting, generating, and processing them one by one. But with AI Agents, you can automate everything—saving hours of work and eliminating human error.

In this video, I break down:
✅ Traditional document workflows vs. AI-powered workflows
✅ 5 key document tasks you can automate today
✅ How businesses are using AI Agents to process thousands of documents in minutes

[Maveen Mushtaq @maveenbhat]

New Tutorial: 🚀 Generate Hundreds of Documents in Seconds with AI!

📺 Watch the tutorial to see how it works! https://t.co/rbi8bsx0sY

With @9one3ai's File Generator Axon, you can:

✅ Upload a template with placeholders
✅ Fill it with AI-extracted data from documents
✅ Generate hundreds of fully completed documents in seconds
No more copy-pasting, no more manual effort. Just upload, extract, and generate—seamlessly.

Want to automate your document generation? Try it out now!

[Maveen Mushtaq @maveenbhat]

🚀 New Tutorial: Automate Document Sorting with AI!

📺 Watch the tutorial here: https://t.co/6J5bi9kgsz

In this tutorial, I walk through how to set up a Classifier Axon and show you how an invoice can follow the invoice route while contracts take a different path—all fully automated!

We’re covering all Axons in this tutorial series—subscribe to our YouTube channel and learn how to automate your document workflows!

[Maveen Mushtaq @maveenbhat]

🚀 New Tutorial Series: Mastering Axons in @9one3ai!

📺 Watch the first tutorial & subscribe for more: https://t.co/AskN7qi6E7

We’re launching a new video series where we break down how to use each Axon in the @9one3ai platform—helping you automate document workflows faster and more efficiently. First up: The Extractor Axon 🔍

[Maveen Mushtaq @maveenbhat]

🚀 Extracting Key Info from Pitch Decks—Automatically!

Watch it in action here: https://t.co/4pwoqvK5ZG

VCs receive hundreds of pitch decks, but manually extracting company names, emails, and contact details takes too much time. With @9one3ai , you can automate this entire process in just minutes.

[Maveen Mushtaq @maveenbhat]

🚀 Instantly Sort Thousands of Documents with AI

Manually organizing documents is time-consuming, repetitive, and prone to errors. But what if AI could do it for you?

With https://t.co/ZsxiRdohUz, you can:
✅ Automatically classify the docs.
✅ Sort thousands of documents into the right folders in minutes
✅ Eliminate manual effort & improve workflow efficiency

Watch the demo to see AI-powered document sorting in action! 📂⚡

[Maveen Mushtaq @maveenbhat]

🚀 Generate Thousands of Termination Letters in Minutes with AI.

Manually creating termination letters? There’s a faster way. With https://t.co/ZsxiRdnK51, you can:

✅ Upload thousands of employment contracts
✅ Automatically generate customized termination letters
✅ Process everything in just two minutes

Watch the video to see how AI-powered document automation saves time and eliminates manual work.

[913.ai @9one3ai]

🚀 Extract Key Information from Hundreds of Contracts in Seconds!

Manually reviewing contracts, invoices, or reports? There’s a better way. With AI Agents on 913ai, you can:

✅ Extract critical details automatically
✅ Classify & route documents instantly
✅ Process hundreds (or thousands) of files in seconds

One of our founders, @ImmoAit , walks through how it works in this quick demo.

📺 Watch the demo here: https://t.co/W2s3oRByrP

This works for any document type, not just contracts. Try it out and say goodbye to manual processing!

[Maveen Mushtaq @maveenbhat]

Google‘s OAuth Verification Process is Broken

If you’re building on @Google Apps, you’ve probably faced this: the OAuth verification nightmare.

For the past seven months, we’ve been trying to get our @Google App verified so users can authenticate with Gmail without seeing that dreaded “Make sure you trust this site” warning. We’re GDPR compliant, working towards ISO compliance, and take security seriously—yet the process has been a bureaucratic maze of endless emails, policy updates, and documentation.

Now, after months of back and forth (and even after Google’s auth team tested our platform themselves), they want another demo. At this point, it feels like an endless loop designed to make it as hard as possible for companies to integrate with Google.

Why is Google making it so difficult for secure, compliant startups to build on their ecosystem? There has to be a better way.

For those who’ve been through this—how did you manage to get verified? Any tips on dealing with Google’s process?

[Naval @naval]

AI is eating UI.