We Almost Launched ArcLab. Then We Started Over.

Akshat Soni — Sat, 04 Jul 2026 19:23:29 +0000

A few weeks ago, I thought ArcLab was ready to launch. The product worked, the landing page looked polished, and I felt like all that was left was pressing the button.

Then I started sharing it with other builders.

The feedback wasn't about bugs or missing features. It was about the experience. A few people said it felt too generic. At first, I wanted to explain every design decision we had made, but after reading the comments a second time, I realized something important. They weren't attacking the product—they were describing how it made them feel.

So we made a difficult decision.

Instead of making small tweaks, we rebuilt it.

We rewrote almost every piece of copy, redesigned entire sections, refined the interactions, removed things we liked, and kept questioning every detail until the experience finally matched the product we wanted to build.

That process taught me something I'll probably carry into every future project.

People don't separate your product from the way you present it. If the first impression feels rushed, they'll assume the product is rushed. If the experience feels generic, they'll assume the product is too.

Today, ArcLab is back.

Not because it's perfect, but because it finally feels intentional. It reflects the level of care we wanted people to experience from day one.

We're getting close to launch, and this time, we're excited to share it.

If you've ever rebuilt something after receiving hard feedback, I'd love to hear your story. What was the comment that changed the direction of your project?

Hello DEV, I’m building ArcLab to help builders ship better AI products

Akshat Soni — Sun, 14 Jun 2026 10:01:38 +0000

Hey DEV 👋

I’m Akshat.

I’m an engineering student and founder-builder currently working on ArcLab.

I joined DEV because I want to document what I’m building, learn from other developers, and share the messy process of going from idea → product → launch.

The product I’m working on is called ArcLab.

The reason I’m building it is simple:

A lot of builders can build fast now.

With AI tools, templates, APIs, and frameworks, it has become easier than ever to create something that looks like a product.

But shipping something useful, safe, and production-ready is still hard.

I kept seeing the same problems:

people building products without validating the market
unclear user flows
auth that does not work properly
API keys exposed in the frontend
products that work locally but break in production
builders launching without knowing if the product solves a real problem

That is the gap I want ArcLab to solve.

ArcLab is a structured execution system for builders who want to go from idea → build → launch with better product and production decisions.

It is not a course.
It is not passive learning.
It is not another random roadmap.

The goal is to help builders make better decisions before they ship.

Right now, ArcLab is in prelaunch.

The full product opens soon, and the first 100 founding seats are $19 before it becomes $49.

I recently got my first prelaunch sale, which is a small milestone, but honestly it means a lot to me. It was the first real signal that someone understood the problem enough to pay for it.

I’ll be sharing what I learn while building ArcLab:

product decisions
landing page changes
launch strategy
technical mistakes
auth/payment/deployment lessons
what works and what fails during prelaunch

I’m still early, so I’d genuinely love feedback from developers here.

If you have built or tried to launch a side project before:

What was the hardest part between having the idea and actually shipping it?

If you want to check what I’m building, here’s ArcLab: https://arclab.systems

Mainly looking for feedback on the positioning and whether the problem feels clear.

DEV Community: Akshat Soni

We Almost Launched ArcLab. Then We Started Over.

Hello DEV, I’m building ArcLab to help builders ship better AI products