DEV Community: Sibirtsev Petr

I built a free mobile dashboard for Lemon Squeezy because I was tired of paying for basic features

Sibirtsev Petr — Sat, 11 Apr 2026 15:15:48 +0000

I built LemonDash because I had a simple problem:

I use Lemon Squeezy for my own store, and I wanted a good mobile app to track what was happening in real time.

I wanted to open my phone and instantly see:

revenue
MRR
sales
subscriptions
refunds
renewals

And I wanted push notifications for every sale.

Sounds basic, right?

But most of the tools I found either:

charged for core features like notifications
did not feel great on mobile
or just were not built the way I wanted

So I decided to build my own.

What I built

LemonDash is a free mobile dashboard for Lemon Squeezy sellers.

It helps store owners track:

MRR
revenue
sales
subscriptions
refunds
renewals

It also sends instant notifications for important store events.

The goal was simple: build something actually useful, mobile-first, and 100% free.

Why I made it free

This was the main thing that bothered me.

Notifications and basic analytics should not feel like premium-only features, especially for indie makers and small SaaS founders.

So I made LemonDash:

free
open source
focused on the features I personally wanted as a seller

I spent my own time and money building it, publishing it, and preparing the launch.

Current status

iOS is live: https://apps.apple.com/us/app/lemondash/id6761753461
Open-source backend: https://github.com/sbrsv/lemondash
Product Hunt launch: https://www.producthunt.com/products/lemondash
Open Launch page: https://open-launch.com/projects/lemondash
Google Play is in review and should be live in about 12 days

What I learned building it

A few things stood out while working on this:

1. “Small pain” products are often worth building

This did not come from a huge startup idea. It came from one annoying problem I personally had.

Those problems are often the best ones to build around.

2. Free can be a real product decision

Sometimes making something free is not just pricing. It is the whole point of the product.

3. Mobile UX matters more than people think

A lot of seller tools work fine on desktop, but feel bad on mobile. That gap is still real.

I’d love feedback

If you are a Lemon Squeezy seller, indie hacker, or micro-SaaS founder, I’d genuinely love to know:

What would you want from a mobile dashboard like this?

And if you’re launching products yourself, I’d also love your feedback on the Product Hunt page.

Voice Cloning for Content Creators That Scales

Sibirtsev Petr — Thu, 26 Mar 2026 10:39:20 +0000

If you publish every day, your voice workflow either helps you grow or slows you down. That is why voice cloning for content creators has moved from a nice extra to a real production advantage, especially for Shorts, TikTok, gaming videos, and faceless channels that live on consistency.

A cloned voice is not just about sounding polished. It is about keeping your narrator identity stable across dozens or hundreds of videos without rerecording every script, fixing every bad take, or waiting on outside talent. For solo creators and small teams, that time adds up fast.

Why voice cloning for content creators works so well

Short-form content rewards repetition. Viewers come back for a format they recognize, and voice is part of that format. If your storytelling channel sounds different every week, or your automation channel switches between random voices, the brand gets weaker.

Voice cloning solves that by giving you a repeatable narrator. You can write a script in the morning, generate the read in seconds, and move straight into editing. That matters even more when you are producing serialized content like gaming recaps, horror stories, anime commentary, or list-style YouTube videos.

It also gives creators more control over pacing. A good clone lets you adjust delivery, tighten awkward lines, and regenerate sections without scrapping the whole take. Traditional recording still has its place, but it is slower and less forgiving when you are on a daily schedule.

Where cloned voices fit in a real creator workflow

The strongest use case is not replacing creativity. It is removing production drag.

If you run a faceless YouTube channel, voice cloning helps you keep one narrator identity across explainers, countdowns, and commentary. If you make TikTok storytelling videos, it helps you maintain the same tone across a series. If you are a gaming creator posting Roblox or Minecraft videos, it can keep your intros, character lines, and recaps consistent without hours in the booth.

The real win is what happens after generation. A voiceover is only useful if it is ready for the timeline. That means clean MP3 export, captions that do not need manual rebuilding, and timing that works with fast edits. For short-form creators, word-level caption alignment can matter almost as much as the voice itself because retention often depends on how readable and punchy the edit feels.

This is where an all-in-one workflow matters more than a flashy demo. A tool that gives you natural speech but leaves you rebuilding subtitles by hand is still costing you time.

What to look for in voice cloning for content creators

Natural sound is the starting point, not the finish line. Most creators should judge a voice cloning tool on four practical questions.

First, does it hold up across volume? A voice that sounds good in one sample can break when you generate ten scripts in a row. Listen for weird phrasing, unstable tone, or lines that sound too stiff for conversational content.

Second, can you move fast? If generation lags, revisions become annoying. The best tools fit the pace of modern publishing, where creators test hooks, swap scripts, and turn around edits quickly.

Third, is the output actually ready to use? MP3 exports are standard, but creators posting short-form should also care about caption support. If the platform can export subtitle files with word-level timing, that saves a surprising amount of editing time.

Fourth, how does it handle safety and ownership? This part gets ignored until it becomes a problem. If you are cloning your own voice, or using voice assets in client work, you need to know the platform takes privacy, consent, and data handling seriously.

The trade-offs creators should understand

Voice cloning is powerful, but it is not magic.

If your original recordings are poor, the clone will usually reflect that. Bad mic quality, inconsistent tone, and noisy samples can lead to a less stable result. Creators who want the best clone should treat the source recording like a real asset, not an afterthought.

There is also a style trade-off. Some creators want highly expressive reads with dramatic swings. Others want clean, controlled narration for automation channels or tutorials. One clone may not be perfect for every format. In some cases, it makes sense to use your clone for the main brand voice and pull from a voice library for character work, comedic cuts, or alternate personas.

And then there is the legal and ethical side. If a platform is vague about consent or data security, that should be a red flag. Commercial creators need policy-first tools because brand trust matters, especially when clients, sponsors, or audiences are involved.

When a standard AI voice is better than a clone

Not every creator needs to clone their own voice right away.

If you are launching a new faceless channel and testing formats, a high-quality stock voice may be the smarter move. It gets you publishing quickly without the setup step of training a custom narrator. This is useful when you are still figuring out your niche, posting style, or script rhythm.

Cloning becomes more valuable once your channel has a clear identity. That is when consistency starts affecting recognition, binge behavior, and brand feel. A custom voice can make your content more repeatable, especially when you are building a content system instead of posting casually.

A lot of creators end up using both. They keep a signature cloned voice for the main narration, then use other voices for skits, alternate characters, dubbed segments, or multilingual experiments.

A faster production setup beats a perfect demo

Creators do not need a voice lab. They need a pipeline.

That is the difference between a tool built for demos and a tool built for publishing. A creator-first platform should help you write, generate, export, caption, and edit without bouncing through five tabs. Speed matters, but speed without control is not useful. You want fast generation plus outputs that slide directly into Premiere Pro, CapCut, Final Cut, or whatever editor you already use.

For creators posting high volume, predictability matters too. Usage-based systems can work well when they are simple and transparent. If one point equals one second of generated speech, you can estimate production costs without guessing.

Vocallab AI is a good example of where the market is going. It is built for fast creator workflows, with near-real-time generation, clone-ready studio tools, and export options that make sense for short-form teams, including MP3 and SRT with karaoke-style word highlighting. That kind of workflow thinking matters more than feature bloat.

How to tell if voice cloning is worth it for your channel

The answer usually comes down to output volume and brand consistency.

If you publish one video every few weeks and enjoy recording manually, voice cloning may be optional. If you publish daily, run multiple channels, create client content, or rely on repeatable narration, it can save a serious amount of time.

It is also worth it if rerecording has become your bottleneck. Many creators do not struggle with scripting or editing. They lose hours to pickups, mic setup, retakes, and fixing lines that sounded fine at first but do not fit the final cut. A solid clone turns those delays into quick revisions.

For agencies and small media teams, the value is even clearer. A shared voice identity can keep a series consistent across editors and publishing schedules. That is hard to do with manual recording unless one person is always available.

Voice cloning for content creators is really about repeatability

The creators who benefit most are not chasing novelty. They are building systems.

A consistent voice helps make a channel feel intentional. Fast generation helps you test more ideas. Ready-to-export assets help you post more often without lowering quality. Put those together, and voice cloning becomes less of a gimmick and more of a production tool.

That does not mean every creator should clone their voice tomorrow. It means the right time to adopt it is usually earlier than people think - especially once posting volume rises and your content starts depending on recognizable delivery.

The best setup is the one that lets you keep publishing without sounding rushed, inconsistent, or generic. If your current workflow keeps getting in the way of that, your next upgrade probably is not another editing trick. It is the voice pipeline behind the content.

I Tested AI Voice Tools for YouTube Automation — What Actually Matters in 2026

Sibirtsev Petr — Tue, 24 Mar 2026 14:37:34 +0000

Recently I started experimenting with AI voice tools while building a small content pipeline.

Nothing fancy — just trying to automate:

short videos
voiceovers
captions

I expected it to be straightforward.

It wasn’t.

🧠 The Gap Between “Top Lists” and Reality

If you Google “best AI voice generator”, you’ll find dozens of articles.

Most of them say the same thing:

ElevenLabs
Murf
Play.ht
Speechify

And yes — they’re all solid.

But after actually using them, I noticed something:

These tools are evaluated like products — not like parts of a workflow.

And that changes everything.

⚙️ What Actually Matters (After Using Them)

When you’re generating one voice clip, everything looks good.

When you try to produce content daily, different problems show up:

generation speed starts to matter a lot
small inconsistencies become noticeable
exporting + editing becomes friction
captions turn into a separate problem

None of this is obvious from feature lists.

🔊 Notes on the Tools I Tried

I won’t repeat specs — just what stood out in practice.

ElevenLabs

Still the most impressive in terms of realism.

But I found myself hesitating to use it for everything:

slower when generating a lot
pricing adds up quickly

Feels more like a premium tool for high-quality pieces.

Murf

Very clean experience.

But it feels more like a presentation / voiceover tool than something you'd plug into a daily content pipeline.

Play.ht

More flexible, especially if you’re building something around it.

But quality wasn’t always consistent for me, depending on the voice.

Speechify

Fast and simple.

But the output felt more functional than expressive — good for utility, less for content.

🤔 Where Things Get Interesting

After trying a few tools, I realized the main issue wasn’t voice quality.

It was this:

Generating voice is easy.
Producing content is not.

Most tools stop at “here’s your audio file”.

But in a real workflow, you still need:

captions
timing
consistency
speed

That’s where friction builds up.

🔍 A Small Observation

While testing, I also came across vocallab.ai.

It didn’t necessarily feel better in every dimension —
but it felt different in what it tries to solve.

Instead of focusing purely on voice generation, it leans more toward:

fast turnaround
built-in captions (SRT)
simple, repeatable workflow

Which made it easier to go from:

“text → usable asset”

without adding extra steps.

📊 Rough Mental Model

After all this, I started grouping tools like this:

Category	Tools	When they make sense
🎯 Quality-first	ElevenLabs	when voice realism is everything
🧰 General tools	Murf, Speechify	occasional usage
⚙️ Flexible / dev-oriented	Play.ht	custom setups
🔁 Workflow-oriented	vocallab.ai	repeatable content production

💡 What I Took Away

The biggest shift for me was this:

Choosing a voice tool is less about “which sounds best”
and more about “what slows you down the least”.

If you’re:

experimenting → most tools are fine
building something real → workflow matters more than features

🤷‍♂️ Still Figuring It Out

I don’t think there’s a perfect tool yet.

Each one solves a different part of the problem.

Curious if anyone here is using AI voice in production —
especially at scale.

What ended up working for you?