Two months ago, I set out to build something in the customer experience (CX) space. For a week I brainstormed problems that both local and online businesses face. One issue in particular stood out.
I came across a business that had been running for more than 15 years but had only a handful of Google reviews. At the same time, newer businesses in the same industry had many more reviews and were ranking higher. That sparked an idea.
---
In the past, I had personally reached out to businesses and manually helped them gather reviews. This time I wanted a solution that was hands-off. The client should be in control and the process should be seamless. That is where SwiftResolve was born.
---
I know how to build and manage websites with WordPress, WooCommerce, or Shopify, but those platforms felt too slow and not suitable for a SaaS product. I wanted something faster and more modern.
While browsing Reddit, I came across three tools: Bolt, Cursor, and Lovable. Lovable seemed like the right choice since I already had enough coding work ahead.
---
I used Lovable’s free daily credits to prompt it to create a SaaS site with basic functionality. By the third day, I upgraded to a paid plan so I could move faster and add more features.
---
Until week 3, I had mostly relied on the free credits, but this week brought the first major roadblocks.
The password reset page refused to work and threw unexplained errors. Fixing it wasted credits and time. Eventually, ChatGPT helped me debug and solve the issue after a few attempts.
The second problem was with Stripe integration. For some reason, Lovable kept overwriting the live key with the test key every time I updated Stripe-related prompts. That was frustrating, but I eventually worked around it.
Other than those two hurdles, the setup went fairly smoothly.
---
At the same time, I began working on the backend. My choice was between Python and n8n, which was gaining popularity at the time. Since I was building an MVP, I went with n8n because it was easier to set up.
Within a few hours spread across three days, the core functions were done. SwiftResolve allowed businesses to request reviews through three channels:
* Email
* SMS (great for local businesses)
* WhatsApp
However, during implementation I learned that WhatsApp required compliance and additional approvals that I did not want to handle at this early stage. Since my goal was to validate the product first, I decided to cut WhatsApp from the live version. If you looked at the backend in the early days, you might have seen it listed.
---
On August 23, I soft launched SwiftResolve on Product Hunt. As expected, not much happened, but it was an important milestone.
By August 29, I migrated the site from Lovable’s servers to another host using GitHub. This gave me more control but also introduced new SEO challenges.
Since the site is built with React, Google was only seeing the page skeleton instead of the actual content. To fix this, I set up prerendering with Cloudflare. It was my first time working with this approach since I had mainly used WordPress before.
The XML sitemap was also broken, so I rebuilt it completely.
---
SwiftResolve now runs on the following stack:
**Frontend Framework & Build Tool**
* React 18.3.1
* Vite
* TypeScript
**Styling & UI**
* Tailwind CSS
* shadcn/ui (built on Radix UI)
* Radix UI
**Routing**
* React Router DOM 6.26.2
**State Management & Data Fetching**
* TanStack React Query 5.83.0
* React Hook Form 7.53.0
**Backend & Database**
* Supabase (authentication, database, edge functions)
* Supabase JS 2.50.5
**Additional Libraries**
* Zod (schema validation)
* Date-fns (date handling)
* Lucide React (icons)
* Recharts (charts and visualization)
* TipTap (rich text editor)
---
To make it easy for businesses to try SwiftResolve without a monthly commitment, I introduced a reduced-price plan. It provides 100 credits per month and includes the same features as the Growth Plan, which costs $ 79 per month.
The next major update will be webhooks. This will allow clients to connect SwiftResolve directly to their CMS instead of uploading data manually or through CSV files.
---
The journey so far has been full of challenges, from debugging errors to learning new tools, but also rewarding. SwiftResolve started as an idea and is now a working SaaS with paying customers on the horizon.
This is only the beginning.
I came across a business that had been running for more than 15 years but had only a handful of Google reviews. At the same time, newer businesses in the same industry had many more reviews and were ranking higher. That sparked an idea.
---
## From Idea to Solution
In the past, I had personally reached out to businesses and manually helped them gather reviews. This time I wanted a solution that was hands-off. The client should be in control and the process should be seamless. That is where SwiftResolve was born.
---
## First Obstacle: Choosing How to Build
I know how to build and manage websites with WordPress, WooCommerce, or Shopify, but those platforms felt too slow and not suitable for a SaaS product. I wanted something faster and more modern.
While browsing Reddit, I came across three tools: Bolt, Cursor, and Lovable. Lovable seemed like the right choice since I already had enough coding work ahead.
---
## Week 2: Getting the Foundation in Place
I used Lovable’s free daily credits to prompt it to create a SaaS site with basic functionality. By the third day, I upgraded to a paid plan so I could move faster and add more features.
---
## Week 3: Early Challenges
Until week 3, I had mostly relied on the free credits, but this week brought the first major roadblocks.
The password reset page refused to work and threw unexplained errors. Fixing it wasted credits and time. Eventually, ChatGPT helped me debug and solve the issue after a few attempts.
The second problem was with Stripe integration. For some reason, Lovable kept overwriting the live key with the test key every time I updated Stripe-related prompts. That was frustrating, but I eventually worked around it.
Other than those two hurdles, the setup went fairly smoothly.
---
## Backend Decisions
At the same time, I began working on the backend. My choice was between Python and n8n, which was gaining popularity at the time. Since I was building an MVP, I went with n8n because it was easier to set up.
Within a few hours spread across three days, the core functions were done. SwiftResolve allowed businesses to request reviews through three channels:
* SMS (great for local businesses)
However, during implementation I learned that WhatsApp required compliance and additional approvals that I did not want to handle at this early stage. Since my goal was to validate the product first, I decided to cut WhatsApp from the live version. If you looked at the backend in the early days, you might have seen it listed.
---
## Soft Launch and Migration
On August 23, I soft launched SwiftResolve on Product Hunt. As expected, not much happened, but it was an important milestone.
By August 29, I migrated the site from Lovable’s servers to another host using GitHub. This gave me more control but also introduced new SEO challenges.
Since the site is built with React, Google was only seeing the page skeleton instead of the actual content. To fix this, I set up prerendering with Cloudflare. It was my first time working with this approach since I had mainly used WordPress before.
The XML sitemap was also broken, so I rebuilt it completely.
---
## The Tech Stack
SwiftResolve now runs on the following stack:
**Frontend Framework & Build Tool**
* React 18.3.1
* Vite
* TypeScript
**Styling & UI**
* Tailwind CSS
* shadcn/ui (built on Radix UI)
* Radix UI
**Routing**
* React Router DOM 6.26.2
**State Management & Data Fetching**
* TanStack React Query 5.83.0
* React Hook Form 7.53.0
**Backend & Database**
* Supabase (authentication, database, edge functions)
* Supabase JS 2.50.5
**Additional Libraries**
* Zod (schema validation)
* Date-fns (date handling)
* Lucide React (icons)
* Recharts (charts and visualization)
* TipTap (rich text editor)
---
## Pricing and Next Steps
To make it easy for businesses to try SwiftResolve without a monthly commitment, I introduced a reduced-price plan. It provides 100 credits per month and includes the same features as the Growth Plan, which costs $ 79 per month.
The next major update will be webhooks. This will allow clients to connect SwiftResolve directly to their CMS instead of uploading data manually or through CSV files.
---
## Looking Ahead
The journey so far has been full of challenges, from debugging errors to learning new tools, but also rewarding. SwiftResolve started as an idea and is now a working SaaS with paying customers on the horizon.
This is only the beginning.
