He shipped AudioPen in 12 hours — now earns $15K/month

Executive overview

Louis Pereira, a part-time indie hacker from India, built the MVP of AudioPen — a voice-to-text AI writing tool — in a single 12-hour hackathon and received unsolicited payments from beta testers the same day. After failing with 15–20 prior products, he treated low-cost experimentation as a deliberate strategy: build many small things, then double down on whatever resonates.

The core insight is that fast shipping combined with building in public creates a feedback loop that turns early testers into paying customers before a polished product even exists.

AudioPen now serves 200,000 users with 5,000+ paying subscribers at $99/year, generating over $15,000 per month as a solo-built, no-code product.

What AudioPen does

• Voice-to-text AI tool that converts spoken thoughts into polished, styled writing
• Supports multiple languages, mobile and desktop; primary use cases are idea capture, meeting notes, and combating writer's block
• Free tier gives limited recording time; paid tier ($99/year or $159/two years) unlocks writing styles, longer recordings, and third-party integrations
• Non-recurring subscription model — users renew on their own terms, no auto-charge

How the MVP was built in 12 hours

• Louis joined a Half Day Build hackathon: participants worldwide raced from idea to revenue between noon and midnight
• Before building, he gathered Pinterest inspiration and designed in Figma — design decisions took longer than expected
• Launched a waitlist with ~10 hours elapsed; early access testers began paying without being asked
• The app was one of five small tools he shipped that week; AudioPen attracted far more engagement than the others

Why it worked from day one

• Timed entry: few comparable voice-to-AI-writing products existed in 2022
• Credibility built from months of building in public on Twitter created a warm, trusting audience
• Community momentum from Half Day Build participants acted as social proof for first-time buyers
• Product hit an unarticulated need — users explained use cases Louis had not anticipated

The playbook for fast shipping

• Build many small experiments you can shut down cleanly, without hurting users
• Design before building — prompting AI to generate UI skips the hard thinking about what the product should look like
• Share designs publicly early and collect signal; spin up an email list the moment resonance appears
• Launch the simplest version that does the core job well, then raise prices incrementally as the product matures
• Act like an indie hacker: personal story and transparency outperform corporate polish when selling to individuals

Staying focused and avoiding distraction

• AudioPen deliberately does one thing well — converting voice to clean text — rather than expanding into adjacent markets
• Louis resisted repeated temptation to chase other app categories that were performing well
• "Slow consistency" over years of refinement is credited as a primary growth driver
• Regular user compliments cite the ability to follow Louis's journey as a builder as a core reason for loyalty

Tech stack and costs

• Bubble for the web app frontend (~$130/month)
• Zeno for backend logic (~$260/month)
• Draftbit for the native mobile app (~$300/year)
• Loop Switches for email (~$800/month)
• Plausible Analytics (~$19/month)
• Variable API costs (transcription and AI) are the largest and most unpredictable line item

Key takeaways

• Repeated failure across 15+ products is not wasted time — it builds skills and increases the probability of eventually hitting product-market fit
• Early signals (unsolicited payments, high engagement relative to similar tools) are worth acting on immediately by expanding and pricing the product
• Indie authenticity is a competitive advantage — users pay partly because they trust and root for the builder as a person