Interview with Gaurav, the Founder Who Built a $69K/Month SaaS by Actually Talking to Customers

Most startups do not die because the founder had a bad idea. They die because someone spent three months polishing an idea that customers quietly did not care about. It is a very expensive way to receive feedback, especially when the feedback is silence.

Gaurav, founder of Fastlane, built an AI-powered launch system for creators that generates short-form marketing content for solo builders, SaaS founders, and mobile app makers. Fastlane reached roughly $69,000 in monthly recurring revenue shortly after its public launch, with more than 1,000 paid users.

The interesting part is not the revenue screenshot. The interesting part is the system behind it. Gaurav and his co-founder Jock made customer conversations part of the product loop, completing around 20 calls a week and roughly 2,000 calls since beginning customer discovery.

That is the whole play. Not a hidden ad funnel. Not a seven-step growth hack wrapped in a $97 Notion template. Calls, notes, patterns, product changes, repeat.

Table of Contents

• 🧯 Why an AI-Powered Launch System for Creators Needs a Real Problem
• 🛠️ The Setup: A Lean Solo Builder Tech Stack
• 📞 The Actual Secret: Twenty Customer Calls Every Week
• 🧠 The Three-Phase Customer Calls Playbook
• 🔀 The UI Pivot: Make Content Decisions Feel Obvious
• 🧱 The Real Talk: Build Less Alone

🧯 Why an AI-Powered Launch System for Creators Needs a Real Problem

What problem was Fastlane built to solve?

Gaurav began building software after seeing independent founders launch small apps. He quickly ran into the same wall many solo builders hit: building is fun, but getting attention is a separate job with its own emotional damage.

Paid ads felt expensive. Short-form content and UGC-style marketing looked promising, but the obvious question kept showing up: what should someone actually post?

That problem became the basis for Fastlane. A user enters a website, Fastlane learns the business and the content style trending in that niche, then creates high volumes of tailored short-form content. From there, users can choose content and schedule it to TikTok, Instagram, and YouTube.

The product is aimed primarily at solopreneurs building SaaS products or mobile apps. In other words, people who need distribution but do not have a content team, a video editor, and an alarming amount of spare time.

Did Fastlane start as a short-form content tool?

No. The original product, called Cassius AI, was much broader. It tried to handle SEO, LLM optimization, Reddit engagement, and short-form content in one horizontal product. Broad tools sound efficient in a planning document. In practice, they tend to become a menu with too many options and not enough reasons to choose any of them.

The team began shipping early versions in July, released initial versions around August, and continued iterating through September. Customer interactions made the pattern clearer: short-form content was the highest-value piece of the product for the people they wanted to help.

So the team narrowed the scope. Development on the vertical short-form content product began in October. A beta reached a waitlist of around 2,000 people in December, and Fastlane launched publicly on March 23.

That is a useful reminder for anyone assembling a no-code startup tool stack or building with AI: the first product does not need to be perfectly focused. It does need to teach the founder what deserves focus.

🛠️ The Setup: A Lean Solo Builder Tech Stack

What technology powers Fastlane?

Fastlane is not a case study in finding the most exotic tools possible. It is a practical stack assembled for speed, with each service doing a specific job:

• Convex for the backend and database.
• Vercel for frontend hosting.
• Railway for the microservice that builds content.
• Clerk for authentication.
• Resend for transactional and mailing-list emails.
• Axiom for logs and observability.
• OpenAI and Claude for copy and captions.
• Fal AI for image and video generation.
• Claude Code to accelerate product development.

The takeaway is not that every founder should copy this exact list. That would be tool cosplay. The real takeaway is that a small team can now assemble an AI funnel builder and ship quickly without waiting for a large engineering department to schedule a meeting about scheduling a meeting.

How did a non-technical co-founder help build the customer intelligence system?

Jock, Fastlane's non-technical co-founder, used Claude Code and connected data sources through MCPs to build parts of the internal system in natural language. That matters because the bottleneck is increasingly less about whether someone can write code from scratch.

The bottleneck is whether someone knows what should be built, what data should be connected, and what question the system should answer. A fast build pointed at the wrong problem is still just a faster mistake.

📞 The Actual Secret: Twenty Customer Calls Every Week

How did Gaurav and Jock get people onto calls?

They made the offer simple. Inside the product UI, they placed a prominent button that invited users to book a call with the team in exchange for seven extra days of access. The button linked directly to Calendly.

It is not revolutionary. That is precisely why it works. The user gets something useful. The founders get context they could not extract from a dashboard. Everyone avoids pretending that a churn chart is a personality trait.

Since August, Gaurav and Jock consistently ran approximately 20 calls per week. Over time, those conversations became a huge advantage. They learned why people signed up, what they were trying to achieve, what confused them, where they discovered the product, and which outcomes actually made them stay.

Why are customer calls more useful than generic marketing advice?

Customer calls are not the marketing strategy by themselves. They are the mechanism for finding the marketing strategy.

When founders understand their best customers, they can identify:

• Where those customers spend time online.
• Which words they use to describe the problem.
• What result they actually want.
• What features feel essential versus merely impressive.
• What acquisition channels brought in people likely to stick around.

Gaurav found that people who signed up simply because they were curious had lower retention and lower product affinity than those who said they needed marketing immediately. That is the kind of difference that gets lost when every signup is treated as equal.

What did the customer intelligence system track?

The team collected AI-generated call notes in Notion, creating a large database of conversation records. They then connected that information to an internal dashboard called Fastlane Customer Intelligence.

The dashboard centralized customer data and made it filterable by fields such as:

• Reason for signing up.
• Business type.
• Company name.
• Subscription duration.
• Usage behavior.
• Customer love score.

The customer love score combined several signals to identify who valued the product most. Product roadmap decisions were then indexed around those people: what were they asking for, what were they achieving, and what made them different from customers who did not stick?

Fastlane also looked for customer success outcomes, including users generating substantial views on TikTok. This shifted the conversation from “Which feature should the founders build?” to “What is helping the most successful customers get results?” That is a much healthier question.

🧠 The Three-Phase Customer Calls Playbook

What does the customer discovery phase look like?

The first phase happens before a founder decides what to build. Gaurav recommends starting with a rough problem space and a rough audience, then reaching out through personal networks and direct messages. Paying someone for their time is also reasonable if that is what it takes to get honest, detailed conversations.

The team followed principles from The Mom Test. The point is to avoid questions that invite polite approval, such as whether someone likes the idea or would hypothetically use a future product. People are generally kind. Kindness is not validation.

Instead, ask about real behavior:

• How are they solving this problem today?
• Can they describe the last time it happened?
• How much time or money did that solution cost?
• What happens if they do nothing?

A painful, repeated problem with real consequences is worth building around. A polite “yeah, that could be cool” is not. It is a trap with good manners.

What happens in usability testing?

Once the team had enough evidence that solo builders needed marketing help, they built an MVP and fought for customers one by one. They searched Reddit threads, posted lead magnets across social platforms, and invited waitlist signups to calls.

During the usability call, the customer received access to the product, shared their screen, and used it while the founder said as little as possible.

This is important. When someone cannot find the right button, the instinct is to rescue them immediately. But that erases the exact evidence needed to improve the product. Silence can be more valuable than a tutorial.

Those usability tests revealed major friction in Fastlane's original interface. The team could see where people got stuck and used that evidence to simplify the experience.

What does customer success research add?

The third phase is about determining whether the product is actually helping customers. Fastlane used PostHog to identify power users, then spoke with those users to understand how the platform fit into their workflow and what outcomes it produced.

The pattern was obvious: short-form content that got views, conversions, and app installs was the core source of value. So the team went all in on it.

This is how an AI-powered launch system for creators should evolve: find the people receiving genuine value, understand the workflow around that value, then remove everything that distracts from it.

🔀 The UI Pivot: Make Content Decisions Feel Obvious

What product change came directly from customer feedback?

Fastlane's refined interface was inspired by Tinder. It places a trending example from the user's niche on one side and Fastlane's equivalent version for the user's product in the center. The user can reject a concept or accept it, then schedule it to a social platform.

Each content idea shows the format that is already trending, such as a slideshow or hook-based video, and adapts that structure to the user's product. Once a user chooses a piece of content, Fastlane generates the caption and allows scheduling to TikTok, Instagram, or YouTube.

The lesson is not “make everything look like Tinder.” The lesson is to use a familiar interaction pattern when customers need to make lots of fast editorial decisions. The team did not guess its way to that decision. They earned confidence in it by seeing users struggle with the old flow.

🧱 The Real Talk: Build Less Alone

What should introverted founders do when customer calls feel uncomfortable?

Gaurav's answer is refreshingly unglamorous: do it anyway. He was nervous before calls too. That does not make the calls less necessary.

Pat Walls makes the sharper version of the point. Founders often avoid calls because they fear customers will say the app is bad. But if customers already think it is bad, avoiding the conversation changes nothing. It just delays the moment when the founder gets useful information.

There is also a broader point here for people trying to launch with AI tools. The current tooling makes it dangerously easy to build in isolation. A founder can now generate code, designs, content, and landing pages faster than ever. Great. That means they can also produce a beautiful wrong answer at record speed.

What should a solo builder test this week?

Set up a direct path to a conversation. Add a booking link to the product, offer a small access extension or helpful incentive, and commit to a weekly number of calls that feels mildly annoying but achievable.

Then keep the loop simple:

1. Record or take structured notes from every call.
2. Store them in one searchable place.
3. Tag recurring pains, requests, outcomes, and objections.
4. Find the customers who love the product most.
5. Build for the patterns they share.
6. Repeat until the product gets simpler and more useful.

Fastlane's growth was not proof that every SaaS can reach the same revenue level on the same timeline. Businesses vary, markets vary, and the internet remains extremely capable of humbling everyone. But the method is portable.

The useful move is not to copy Fastlane feature for feature. It is to remix the operating system: identify a painful problem, talk to the people living with it, build the smallest useful response, and treat every customer interaction as evidence. That is the AI-powered launch system for creators worth borrowing.

❓ FAQ

How many customer calls did Fastlane complete?

Gaurav and Jock completed around 20 customer calls per week and roughly 2,000 calls since they began customer discovery.

What is Fastlane?

Fastlane is an AI-powered short-form marketing tool for solo builders, SaaS founders, and mobile app creators. It generates tailored content ideas and helps schedule approved posts to TikTok, Instagram, and YouTube.

What questions should founders ask in customer discovery calls?

Ask about past behavior rather than hypothetical interest. Useful prompts include how the person currently solves the problem, the last time they faced it, what it costs them, and what happens if they ignore it.

Why should founders stay quiet during usability tests?

Speaking too quickly hides product friction. Letting customers navigate independently reveals where the UI, onboarding, or language is confusing, which gives the founder concrete evidence for product improvements.