A bit of background: I'm a Master's student at SRH Heidelberg
writing my thesis on autonomous database tuning using Bayesian
optimization. In parallel, I built two full-stack SaaS platforms
from scratch and went through the full cycle of building,
auditing, fixing, and trying to sell them.
Here's what I actually built and what I learned.
Project 1: Bookify — Multi-Tenant Appointment Booking SaaS
Stack: Node.js 18 / Express 5.1 / React 18.2 / MongoDB /
Redis / Socket.io / Stripe / BullMQ
What it does: full multi-tenant booking platform — think
Calendly but white-labeled per business, with 5-role RBAC,
16-currency Stripe payments, embeddable booking widgets,
waitlist management, and 8 languages including Arabic RTL.
What I'm proud of: 940 passing tests (328 backend, 612
frontend, 100% pass rate). The CI/CD pipeline took 3 days to
get fully green due to Node version mismatches, V8 OOM during
minification, and Docker Compose V2 syntax changes.
What hurt: I submitted to CodeCanyon and got hard-rejected
in under 1 hour with a generic message on both projects
simultaneously. No specific feedback. Opened a support ticket
— got a response 3 days later saying "Envato is competitive."
Not helpful.
🔗 Bookify on Gumroad (€99)
Project 2: DriveHive — Full-Stack Car Rental Platform
Stack: Spring Boot 3.2.4 / Java 21 / React 19 / MongoDB /
Redis / Neo4j (optional) / Stripe / Leaflet GPS tracking
What it does: full car rental system — vehicle browsing,
bulk bookings for corporate fleets, live GPS tracking, EV
charging station locator, subscription tiers
(FREE/PREMIUM/ENTERPRISE), custom report builder with
PDF/CSV/Excel export, and a 10-module admin panel.
What I learned building this in Java after years of Node:
Spring Boot's dependency injection is incredibly clean for
large codebases but the cold start time on free hosting tiers
is brutal (60-90 seconds). Not a great demo experience.
🔗 DriveHive on Gumroad (€79)
The CodeCanyon experience
Both got hard-rejected in under 1 hour. Identical generic
messages. After 3 days the support team replied with
essentially "we're competitive, good luck."
My read: CodeCanyon is in structural decline post-Shutterstock
acquisition. They're restricting new sellers to protect
existing catalogue revenue. If you're building source code
products, go directly to Gumroad and Codester. Don't waste
time on Envato.
Honest lessons
1. A green CI pipeline matters more than you think.
Buyers who actually evaluate source code run the tests first.
I have 940 passing tests. Most competing templates have none.
2. Documentation is underrated.
I have 7 documentation files for Bookify and 4 for DriveHive.
Most competing templates ship a single README. Thorough docs
signal a professional product.
3. Multi-tenant architecture is genuinely hard to get right.
Complete data isolation per business via tenant middleware,
plan-gating enforcement, 5-role RBAC — this took months to
design correctly. It's the core differentiator of Bookify.
4. Building while writing a thesis is doable but brutal.
Bayesian optimization by day, JWT refresh token rotation by
night. Context switching between academic research and
production code is its own skill.
5. Free hosting tiers destroy demo experiences.
Spring Boot apps with 60-90 second cold starts on Render's
free tier look broken to potential buyers. Pay the $7/month
for a warm server if your demo is part of your sales process.
If you're building something similar or evaluating source code
platforms to buy rather than build — happy to answer anything
in the comments. Also genuinely curious if anyone else has
dealt with CodeCanyon rejections and found a better path.
Top comments (0)