DinexSpace is a web platform for MetaTrader 5 Expert Advisors (EAs, trading bots), with EA rental, custom EA development and an MT5 indicator library in one place.
The problem
Traders want to see how an EA was tested before renting it, and traders with their own rules need them turned into a working EA.
Journey
- 26 Jun 2026: first commit, an EA marketplace with an affiliate system and an admin back office.
- 28 Aug 2026: licence keys for MT5 EAs, checked with HMAC.
- 18 Sep 2026: backtest figures published with a small-sample warning, plus SEO work.
- 11–12 Oct 2026: five tagged deployments, then a review of customer-facing content.
My role
Nguyễn Hải Minh designed and built the system together with AI, developing with Claude and Codex and coordinating several AI agents.
Time spent
About 83 hours, estimated from commit history, not a timesheet, across 38 working sessions. Research and tester runs that left no commit are not counted, and many commits were made by a team.
Tools and stack
- Next.js frontend
- Python and FastAPI backend
- Docker
- Google sign-in
- Multi-layer automated testing
- A Telegram bot for operations
Checked
- The public pages open (screenshots of 10 Nov 2026).
- The DinexSpace seller page on MQL5 Market listed 31 products on 10 Nov 2026.
Not checked
- Trading results of the EAs. This page makes no claim about profit.
- Visitor and customer numbers are not published here.
What can be observed
EAs for monthly rent with backtest reports, a four-step custom EA process, a 14-page indicator library with 8 sample .mq5 files, and a test lab page.
FAQ
What is DinexSpace?
DinexSpace is a MetaTrader 5 EA platform at dinexspace.com with EA rental, custom EA development and an indicator library.
What was checked on DinexSpace?
The public pages open, and the DinexSpace seller page on MQL5 Market listed 31 products on 10 Nov 2026. Trading results were not checked.
What is DinexSpace built with?
Next.js, Python with FastAPI, Docker, Google sign-in and a Telegram operations bot.
Who built DinexSpace and how long did it take?
Nguyễn Hải Minh designed and built it with AI tools, in about 83 hours across 38 sessions (estimated from commit history, not a timesheet).