A human in chat.
Not an AI bot.

Most B2B SaaS products built support around the same model: automated AI chatbot → if it can't help, it opens a ticket → you wait 24-48 hours → a junior rep replies with canned answers → if an engineer is needed, another 2-3 days to escalate. Net result — a week to solve a question that in reality would take 5 minutes with the right person. We didn't want that kind of support, so we built it differently.

Why tier-0 AI + tier-1 junior doesn't work

The classic SaaS support ladder:

The problem: your technical question often needs tier-3, but it lands at tier-1. Junior asks for repro, sends screenshots, drives you through FAQ. By the time the question reaches engineering, a week has passed.

One chat, correct routing

You message @tgon_support_bot on Telegram. From there, routing depends on the question:

• Billing / license / activation → product-ops (usually Alina). Replies immediately.
• UI / workflow / how-to → frontend engineer (Maxim). Figures it out fast.
• Sender / proxy / anti-ban → core engineer (Ilya). Knows MTProto by heart.
• AI / LLM / Qualifier → AI lead (Artem). Built the AI modules himself.
• Custom integration / API → CTO (Nikita). For Agency customers.

The routing happens on the support manager's first reply: they look at the gist and forward to the right chat. You don't see this "kitchen" — the right person just replies.

Who replies — with names

Plus the CTO (Nikita) for custom integrations, and the founding engineer (Denis) — senior on hard cases. 5-7 people, all real engineers, all have written code in the product.

Response time — honestly, by tier

No "guaranteed 15 minutes 24/7" — we're a small team, and such promises would be lies. Realistic numbers:

• Starter ($49/mo): community queue during business hours (10:00-20:00 Moscow time). Usually 2-4 hours. Weekends — next business day.
• Pro ($89/mo): priority queue. 1-2 hours during business hours. Critical bugs — faster.
• Agency ($169/mo): dedicated channel. Response in 30-60 minutes. Dedicated solution engineer for custom work.

What we do NOT do

• No phone calls. Everything stays in chat — text is searchable later, and async works better for both sides.
• No ghost replies. No vague "a colleague will get back to you." If we hand off, we write "passing to Ilya — he'll reply within an hour."
• No video calls for simple issues. A screenshot + description is usually enough. Zoom is reserved for onboarding or custom integration work.
• No auto-closing tickets. If you don't reply for 3 days — we don't auto-close. We wait for you.
• No satisfaction surveys. You already got your answer — don't waste time on 5-star forms.
• We don't force you to read the FAQ. If the question is simple — we reply, not redirect to docs. Docs exist as a backup for those who prefer reading.

Support is part of the product. We deliver it the way we'd want to receive it from other SaaS companies.