Building in the open · pre-launch

Telegram → MT4/MT5 copy trading,
rebuilt by someone tired of the gaps.

Today's options charge separately for the copier and for MetaAPI — the service that actually runs MT4/5. They silently drop signals their parser can't read, and treat every signal channel as equally trustworthy. Signalrelay runs MT4/5 itself, surfaces the signals the parser couldn't handle, and scores channels on what they actually do. Signal channel subscriptions stay yours. It's not ready yet. The waitlist gets the beta when it is.

→ Get notified when beta opens
✓ You're on the list. I'll email when the beta opens — and roughly once a month before then with build updates.

No spam. No "limited time" nonsense. ~1 email every 4–6 weeks until launch. Reply with feedback anytime — every email goes to me directly.

Status In development
Beta target Late 2026
Built by Solo engineer
Stack Self-hosted MT4/5
Why this exists

Three things nobody in this space gets right.

I'm a software engineer with twelve years of experience. I started using Telegram → MT4 copy services as a retail trader and got progressively more annoyed at the same three problems showing up everywhere. None of them are technically hard. They just don't get fixed because the incumbents have no incentive to fix them.

The MetaAPI dependency. Most "Telegram-to-MT4" services don't actually run MT4 themselves — they require you to subscribe separately to MetaAPI to do it. Two services, two bills, two configurations, two places where your broker credentials live. It's not technical necessity; it's outsourced complexity. Signalrelay runs MT4/5 directly. One subscription covers both. Your signal channel subscriptions stay yours, as they should — that's a relationship between you and the channel, and Signalrelay shouldn't be in the middle of it.

The silent parser. When a signal channel posts something the parser doesn't understand — slightly unusual format, emoji where it expects a colon, multiple TPs in an unexpected order — most services just... drop it. You don't see what failed. You don't see what was ambiguous. You just see fewer trades than the channel actually sent. The parser should tell you when it's confused. Signalrelay's does.

The trust black box. Every signal channel claims a 90%+ win rate. Most of those numbers are cherry-picked, cropped to good months, or based on demo accounts. No copy-trading service tells you which channels have actually been reliable, which ones use martingale that's one bad streak from blowing up, which ones delete losing trades from history. Signalrelay scores channels on what they actually do, not what they say.

Varinder  ·  Software engineer · Brampton, ON  ·  Follow the build
What's being built

Honest about what's done and what isn't.

Most "feature lists" on pre-launch pages are wish-lists dressed as facts. Here's what's actually shipped, what I'm building now, and what's on the plan. I'll update this page as things move.

Building now

One subscription. MT4/5 included, MetaAPI not required.

Signalrelay hosts MT4 and MT5 terminals directly. You connect your broker once; we run it 24/7 on isolated infrastructure. No separate MetaAPI subscription, no second monthly bill, no separate vendor relationship to maintain.

why · the MetaAPI subscription requirement is a UX problem, not a technical one. fixing it is the table-stakes differentiator.
Building now

Parser that tells you when it's confused.

Regex-first for known formats, fallback to a small local model for the long tail, dead-letter queue for the rest. When a message can't be parsed cleanly, you see the message, you see why it failed, and you decide what to do. No silent drops.

why · the difference between "this service is reliable" and "this service quietly misses 15% of signals" is whether the parser surfaces its failures.
Designed · building soon

Signal channel trust scoring.

Every channel scored on observable behaviour: signal volume consistency, win rate measured against actual broker fills (not the channel's own claim), drawdown patterns, signal-deletion frequency. You see the score before you copy.

why · the entire industry is downstream of signal channels. evaluating them objectively should be table stakes; nobody does it.
Designed · building soon

Per-channel risk rules.

Aggressive sizing on the channel you trust; conservative on the new one you're testing. Per-channel max exposure, per-channel TP behaviour, per-channel symbol mapping for brokers that name things differently. Your capital, your rules.

why · trusting all signal channels equally is how accounts blow up. risk controls have to be per-channel, not global.
Planned for beta

Real observability for your trades.

Public status page, per-trade execution logs, latency distributions you can see — not a number on a marketing page. When something doesn't execute the way you expected, you have the trace to figure out why.

why · trading-tool reliability is about transparency more than metrics. you should be able to debug your own service.
Under consideration

Open-source signal parser library.

The parts of the signal parser that aren't a competitive differentiator will likely be released under a permissive license post-beta. The full scope and license are still being thought through — but the intent is to make the signal-extraction logic inspectable for anyone who wants to verify what's happening with their messages.

why · trust in this niche has to be earnable. inspectability of the parser is one of the strongest unfakeable trust signals.
Follow along

I'm building this in the open.

Weekly progress, technical deep-dives on what's hard about this space, and honest notes when things break. Pick whichever you check most.

Common questions

What you're probably wondering.

When does the beta open?

Targeting late 2026. I'll give everyone on the waitlist at least two weeks' notice, with detailed setup instructions and what to expect. The beta will start small — probably 15–25 people — to make sure the experience is solid before opening wider.

What does joining the waitlist actually do?

You get an email when the beta opens, and roughly one update email every 4–6 weeks before then. That's it. No marketing campaigns, no "act now" pressure, no upsells. If you reply, you'll be replying to me directly — every email comes from a real human.

Will it be free?

The closed beta will be free in exchange for honest feedback. Public pricing isn't finalised yet — I'd rather price it once I know what real users find useful. Current plan is a single tier (no MetaAPI add-on, no surprise fees) in the $25–50/month range. The waitlist email will have the final pricing two weeks before launch.

Who's building this?

Me — Varinder, software engineer with twelve years of experience in distributed systems and trading-adjacent infrastructure. I'm based in Brampton, Ontario. I'm a real person and I show up under my own name; you can find me on X, GitHub, and LinkedIn from the links in the footer.

This isn't an anonymous offshore operation. Most copy-trading services in this space don't tell you who's running them. That's a deliberate signal you should pay attention to when evaluating any tool that's going to hold your broker credentials.

Why should I trust you?

You shouldn't, yet. That's literally why this page is a waitlist instead of a "start free trial" button. Trust takes time to earn and there's no shortcut. What I can do is build openly: ship visible progress, write honestly about what's hard and what breaks, open-source the parts that don't need to be proprietary, and let you decide based on a year of consistent work whether I've earned it.

Follow the build (links above). Read what I write. Talk to early beta users when they exist. Then decide.

How is this different from TelegramFXCopier / TelegramCopyTrader / etc.?

The three core differences: (1) bundled pricing — MT4/5 hosting and the Telegram session pool are included, you don't subscribe to MetaAPI separately or rent a VPS; (2) signal channel trust scoring as a first-class feature, not a marketing afterthought; (3) parser transparency — when a signal can't be parsed cleanly, you see what failed instead of just seeing fewer trades.

The incumbents have ~5 years of head start, more features today, and existing customers. What they don't have is structural reason to bundle pricing or surface parser failures, because both would reduce ARPU or expose limitations. The differences below are positioning differences, not feature races.

What if I want to chat before the beta opens?

Email me at hello@signalrelay.io. I read everything; reply rate is high. If you're a Forex trader, a signal channel operator, or someone who's been burned by an existing copy service and has opinions on what should be different — I'd genuinely like to hear from you.

Want to see how this turns out?

The waitlist gets the beta first, monthly build updates, and the final pricing two weeks before public launch. One field, no commitment.

✓ You're on the list. Talk soon.