Navigating the MVP Journey: Building Deescalator One Step at a Time

Building the MVP for Deescalator has been a long, scrappy, and rewarding project. As a solo developer with limited time and resources, every decision has been about making progress without getting stuck chasing perfection. From choosing the right tools to dealing with unexpected roadblocks, this journey has taught me a lot — and the product is better for it. Here’s a look behind the scenes.

Key Challenges Along the Way

Tool Selection

The first step was picking tools that were reliable, free (or nearly free), and wouldn’t slow me down: • MongoDB: Flexible, easy to work with, and a good fit for the early data model. • Google Cloud: Hosting backend scripts and services for less than a dollar a month. • Vercel: Handling frontend deployments with minimal friction. • Auth0: Managing user authentication securely, on the free tier for now.

Each tool had its quirks, but overall they allowed me to build fast without worrying too much about infrastructure.

Database Design

I kept the database schema as simple as possible to start using Deescalator quickly myself. Instead of perfecting everything upfront, I added and adjusted fields as real needs appeared. This flexible approach fit well with MongoDB’s document model and helped me keep momentum.

Authentication Setup

Getting authentication right was more painful than I expected. It was my first time integrating a third-party SSO like Auth0 alone, and debugging authentication flows took more time than planned. But it was a necessary step to make sure user data is properly secured from the start.

Performance

Since I’m staying on free hosting tiers for now, performance mattered more than ever. I optimized frontend loading with React and Vite, kept backend queries lean, and made sure database usage stayed efficient. Continuous small improvements have made a noticeable difference in speed and responsiveness.

Product Focus

I resisted the temptation to add every feature I could think of. The priority was simple: make logging conflicts easy and useful. I added just enough extra functionality — security improvements, a few quick wins for user experience — to make the MVP worth testing.

UI Decisions

I deliberately kept the UI simple and practical. I used Mantine’s standard components to avoid wasting time on custom designs, and focused instead on making sure the app works well on both desktop and mobile. There’s room to polish things later — but for now, usability beats beauty.

Lessons Learned

The biggest lesson? Pragmatism wins. With limited time, it’s better to ship something functional and learn from it than to build a “perfect” app that no one ever uses.

Working on Deescalator has been demanding, but it’s also been the most fun and instructive side project I’ve done so far.

What’s Next

The MVP is almost ready for a broader release. Using Deescalator myself over the past six months has already helped me tweak and improve it. I’m planning to keep using and improving it no matter what — but whether I invest more time, money, and energy will depend on how the public launch goes.

Thanks for reading, and stay tuned for more updates!