Built a live SaaS product in 55 days with zero coding background

VIDEO SUBMISSION:

https://youtu.be/oHPTGZ-zbYA

https://drive.google.com/file/d/1X6HogSQFXsYdX9eZLinWv-w6qHTwXwxw/view?usp=sharing

ONE-PAGER & PRESENTATION: attached below

RELEVANT URLS:

Landing page: https://moveincircle.com

Product: https://app.moveincircle.com

ACHIEVEMENTS:

Circle by MoveInCircle is a fully functional, live SaaS product that collects photos from event guests and turns them into finished digital keepsake albums. It is live at moveincircle.com and app.moveincircle.com with custom domains, real users, and a complete feature set.

I built this as a non-technical founder with zero coding background. Six months ago, I did not know what a code repository was. Today, Circle includes: event creation with cover photos and date ranges, QR code and shareable link generation, guest uploads with no app or signup required, upload moderation (approve/reject/restore), PDF keepsake album generation with cover page and contributor names and captions, ZIP download of all media, free and paid tiers with upgrade codes, full event lifecycle management (Draft, Live, Closed, Reopen), magic link authentication, and a branded landing page.

The product went from idea to live in 55 days (December 20, 2025 to February 13, 2026). Total cost: under $1,000 in AI tool subscriptions. Traditional development cost for the same product: $50,000 to $100,000.

TOOLS:

My vibe-coding stack evolved throughout the project:

Special Mention about Wispr Flow - Changes my life and I hardly type anymore.

Phase 1: ChatGPT as the brain and advisor, Replit for backend (copy-pasting ChatGPT code), Lovable for frontend. Problem: none of the tools were talking to each other. Constant context loss.

Phase 2: After Sarah and Tyler's class on VS Code and Claude Code, I migrated the backend from Replit into VS Code with Claude Code. This was the first breakthrough. I could now develop in one environment with AI assistance built in.

Phase 3: Moved hosting from Replit to Railway. Set up Postgres database, persistent file storage, and custom domains on Railway. This solved the hosting problems.

Phase 4: The biggest breakthrough. I realised Claude Code could control the frontend too, not just the backend. Moved Lovable frontend editing into VS Code via GitHub sync. Then the landing page too. Claude Code now controls backend, frontend, and landing page. Learned Git and GitHub from scratch during this project.

Current stack:

VS Code with Claude Code: controls backend, frontend, and landing page. Built the entire API, PDF generation, authentication, file handling, and event lifecycle.

Lovable: frontend deployment platform, with code edited through VS Code and pushed via GitHub.

Railway: cloud hosting with Postgres database, persistent file storage, and custom domain configuration.

GitHub: two private repositories (backend and frontend), full version control.

Claude (chat): strategic planning, market research, documentation, and decision support.

METHODOLOGY:

I used a multi-AI workflow where each tool had a specific role. Claude Code was the builder. Lovable was the frontend canvas. Railway was the infrastructure. Claude chat was the strategist. When I hit a wall with one tool, I used another to solve it.

I recorded every decision in a Decisions Log and Parking Lot document to prevent scope creep. Features were either "build before deadline" or "parked for post-submission." This discipline kept the project on track.

I iterated rapidly: idea to feature in hours, not sprint cycles. When something broke, Claude Code fixed it. When a design needed changing, I pushed through VS Code to Lovable via GitHub. The entire development cycle was compressed because AI handled the code while I handled the product decisions.

LEARNINGS:

1. Context is everything. The biggest productivity loss early on was re-explaining the project to AI tools. Once I consolidated into VS Code with Claude Code, development speed increased dramatically.

2. AI tools are not a replacement for product thinking. The tools write code, but deciding what to build, what to defer, and what to cut requires human judgment.

3. Git is not optional. Learning version control was one of the most valuable things I did. It gave me confidence to experiment because I could always roll back.

4. Constraints accelerate progress. The 55-day timeline and the decision to park features forced focus on what matters: a working product that solves a real problem.

5. The biggest risk for a non-technical founder is not the code. It is scope creep. A clear decisions log saved this project multiple times.

ATTACHMENTS (list of files to upload):

1. Video: Vibe Coding Builder Submission You tube, and G Drive links attached for the same video.

2. Presentation Deck of the Journey and Product

3. One-pager Summary on Circle