Speasy: Automated audio daily podcasts

speasy.app/

Speasy

Live prototype of the app I careated

Introduction

Speasy started as a personal productivity hack: I wanted a frictionless way to turn newsletters and articles into high-quality audio I could listen to during commutes, workouts, or downtime. What began as a weekend proof-of-concept in Replit evolved into a polished SaaS product—integrated with Supabase, Stripe, and AI-driven summarisation—that automates the entire pipeline from newsletter → digest → inbox → podcast feed, without me lifting a finger.

The Problem: Drowning in Unread Content

Like many busy professionals, I was drowning in unread newsletters and articles. My reading list felt endless, and the guilt of “I’ll read it later” was real. Staying informed is critical, but the modern pace of work leaves little time for deep reading.

This frustration became the inciting incident: what if I could reclaim those wasted moments—commutes, chores, idle time—and turn them into learning time?

Early Proof of Concept

• Replit Build: I hacked together a simple web interface that converted URLs into text and fed it through OpenAI’s TTS, syncing output to Overcast as a private podcast feed.
• No-Code AI: Inspired by no-code AI workflows, I used tools like Make and Claude Workbench to refine prompts that summarised content into concise, listenable scripts.
• Iterations: From share-sheet shortcuts, to email services and input fields, there have been many flavours of the idea.
• Automatic content scraping: chunking, TTS conversion, merging via FFmpeg.
• RSS feed generation, and playback interface.
• The result: my first automated “read it later” pipeline—raw, but working.

Scaling Into a Real Product

Rebuilding in v0 and Cursor w/ Claude Code

I migrated from Replit to v0 for UI iteration, then moved into Cursor for structured development. Core integrations followed:

• Supabase for storage, authentication, and RSS feed generation.
• Stripe for subscription management and payments.
• Resend to power onboarding and weekly digest emails.
• Inoreader to automate ingesting top newsletters and articles across design, tech, and business.

From Articles to Audio

Speasy evolved beyond simple URL-to-audio:

1. Summarisation Prompts: Generate bullet-point summaries and key takeaways.
2. AI TTS: Convert these summaries into high-quality speech using OpenAI’s TTS.
3. Digest Creation: Combine weekly articles into 8–10 minute category-based audio digests.
4. Private Podcast Feeds: Deliver everything via personalised RSS feeds compatible with Apple Podcasts, or Overcast.

Recruiting Beta Users

I invited early adopters via LinkedIn:

• Collected detailed feedback on features, UI, and workflows.
• Validated demand for condensed audio, playlist control, and multiple feeds.
• Iterated based on insights: better summarisation, “demo mode” for frictionless onboarding, and a redesigned dashboard

Design & Visual Evolution

As the product matured:

• I uplifted the dashboard’s visual style, improving hierarchy and readability.
• Added blog content for SEO and education.
• Built a demo version allowing non-registered users to stream real digests—reducing signup friction.

Automation in Action

Today, Speasy runs end-to-end with minimal intervention:

• Newsletters flow in automatically via RSS.
• AI summarises, converts, and assembles weekly digests.
• Personalised podcast feeds update instantly.
• Resend emails users their weekly digest.

From newsletter → digest → inbox → podcast feed happens without me lifting a finger. It’s the kind of automation I wish existed years ago—fewer tools to check, more time reclaimed.

Testing infrastructure

The codebase has extensive testing (397 tests passing) including:

• Visual regression testing
• Accessibility testing
• Performance benchmarking
• Security monitoring system
  • CSRF protection middleware testing
  • API integration testing
  • Structured logging (replacing console statements)
  • Quality gates and pre-commit hooks

Sophisticated Analytics & Monitoring

• Performance monitoring utilities
• Bundle size optimisation
• Core Web Vitals tracking
• User behaviour analytics

Technical Debt Prevention System

• Automated quality gates
• Performance budgets
• Dependency security monitoring
• Import path consistency enforcement

What’s Next

I’m exploring:

• Deeper personalisation (AI-driven playlisting and topic exploration).
• Mobile-first experience for seamless consumption.
• Learning agents: turning passive listening into active growth with deep-dive follow on questions.
• Using React Native Expo mobile app to expand to new platforms

Takeaway

Speasy reflects how far rapid AI prototyping can go when mixed with product thinking. From Replit hacks to a fully automated SaaS with real paying users, it’s proof of what’s possible when design, AI, and automation converge.

Visit Speasy