MeerGuard VPN

Problem

Client work is where the team picked the stack up: Xray, VLESS, server configuration. Our own configs for friends came next, and constant tuning with them — in Russia far from every setup works reliably, and the ones that do had to be found rather than looked up. That is what grew into the idea of a subscription service. Instead of committing to an iOS app first (slow review, expensive development), we started with a Telegram bot to validate demand fast, then expanded to a Mini App, website and iOS client once the hypothesis held.

Solution

We built a multi-platform product around a single backend: Telegram bot as the main entry point, Telegram Mini App and website for in-app management, iOS app for tunnel control. The bot handles the full subscription lifecycle: registration, plan selection, payment, key generation, server management. Four payment providers integrated (YooKassa for Russian cards, CryptoBot for crypto, Telegram Stars, Tribute). Backend on Python/Aiogram 3 with PostgreSQL, Redis and Celery for background tasks. 5 Docker containers in production. Admin panel inside the same bot: user management, broadcast messaging, promo codes (single and bulk), refund processing, payment analytics by provider. Referral system with two-sided bonuses. Integration with Remnawave panel via REST API for VPN server management across 8 countries.

Results

• +9,113 users in 10 months, 1,045 subscriptions active right now
• +1,185 of them arrived through referrals from existing users
• +39% of signups convert to a payment — 7,062 successful payments
• +4 client surfaces: Telegram bot, Mini App, website, iOS app
• +Breakeven by month 3. Subscriptions: ~$2/month, ~$5/3 months
• +Full test coverage, clean architecture

Role & timeline

Started with the bot and the VPN setup; now on me: YouTube promotion, management, advertising and tech decisions. Alongside: a backend developer (bot + backend), a DevOps engineer (VPN configuration), a full-stack developer (iOS, Android, Web, Mac)