🎧 From hacky prototype to personal podcast app – The speasy journey so far
Speasy turns saved articles into personal podcasts. Built with AI-first tools, it's designed for busy professionals who want to learn on the go—hands-free.

Richard Simms
· 5 views

6 months ago, I hacked together a proof-of-concept on Replit for an idea that wouldn’t leave me alone. I was drowning in saved articles I never had time to read, and I wondered: what if I could listen to them instead? That weekend on Replit, I vibe coded to turn an article into audio. It wasn’t pretty, but it worked well enough to share with a couple of friends.
Their reaction gave me that spark of validation. They loved the concept of turning “read later” into “listen now,” but they also pointed out a dozen ways it could be better. Over the following weeks, I found myself in a cycle: build, get feedback, iterate, repeat.
For example, one early user wanted more natural-sounding voices. Another needed an easier way to add articles. Every piece of feedback nudged Speasy in a better direction, one small tweak at a time.
Being a solo builder, I decided to take an AI-first approach to development – basically using every tool I could to cover my weaknesses. I’m not a one-person coding army, so I leaned on some outstanding (and often AI-powered) tools to help bring Speasy to life:
- Replit – for quickly prototyping the idea in a live environment.
- v0 – to generate the initial UI so I didn’t get bogged down in design.
- Cursor – to debug and refactor code faster than I could alone.
- Devin (an autonomous AI agent) – to run some coding tasks in the background, like a 24/7 pair programmer.
- Supabase – to handle the backend and database without me writing it all from scratch.
- OpenAI – for the smarts behind the scenes, parsing content and making the experience more intelligent.
- Stripe – for easy payment integration (gotta prep for sustainability!).
- ElevenLabs – to turn text into realistic speech, giving each article its own natural voice.
Using these tools felt like having an extended team of specialists, even though it’s really just me here. This approach let me sidestep a lot of my coding pitfalls and focus on the product itself. It’s humbling (and a little wild) to see how far an idea can go when you let AI handle the parts you’re not great at.
The result of all this: Speasy evolved into an app that helps people turn their saved articles into personalised podcast episodes. In practical terms, that means you can queue up the articles you’ve been meaning to read and listen to them in a friendly voice while you commute, exercise, or cook dinner. No more staring at a screen or feeling guilty about that backlog of content – now you can learn on the go, on your terms.
Today, I’m coming full circle with our early adopters. I’ve started looping back with the first beta users (the patient folks who saw that raw Replit demo months ago) and sharing the new-and-improved version of Speasy. We’re exploring it together to see what’s working and what still needs love. It’s exciting to be in co-creation mode with them – listening to their experiences, brainstorming fixes, and dreaming up new ways to solve their jobs to be done side by side.
This journey has been anything but linear, and there’s plenty left to build. But taking a step back, I’m feeling grateful. Grateful for the early testers who believed in the idea enough to give honest feedback. Grateful for the AI tools that helped a solo founder move a little faster. And most of all, grateful to be turning a personal pain point into something that might just help others, too. It’s a slow, steady grind, but I’m loving the process.
I’m not sure exactly where this leads yet, but I know I’m building something I believe in—with the people it’s meant to serve.