The CRM for crypto runs on Telegram

Let's be honest — nobody's closing a round over email. The intro, the terms, the "gm, still on for Tuesday?" — it all happens in Telegram, usually in a group chat with twenty people and zero structure. That is not a quirk of one cycle. Telegram passed a billion monthly users in 2025, its sister network TON raised $400 million from crypto VCs the same week, and every exchange, fund, market maker, and project you have ever dealt with has a Telegram group with your name in it.

Normal CRMs assume your relationships live in a tidy inbox and a neat pipeline. Yours don't. They live in a thousand unread threads, and a little context dies every time you scroll past one. The channel is not the problem — Telegram is where the deals are. The problem is that nothing is writing it down. Pantheon makes Telegram the place your deals get remembered instead of the place they quietly vanish.

A market-making desk's BD does not run through a pipeline tool. It runs through a specific kind of Telegram group: the desk, the project's founders, and often the exchange's listing team, in one chat. The first message is an intro from a VC or an advisor. Then it's the term conversation — a loan-and-call-option structure or a retainer, the spread and depth KPIs, which venues. Then the listing date moves twice. Then a different founder joins the group and the original one goes quiet.

What breaks: the desk is in fifty of these groups at once, and each one has a slightly different set of people, half of whom have no username. Two months later a founder DMs from a new handle — "we spoke about liquidity for the TGE" — and the BD lead has to scroll three groups to work out which project and which terms. The advice to projects is already "don't outsource the decision to a deck and a Telegram intro"; the desk on the other side of that intro needs to remember it just as precisely.

What a Telegram-first CRM has to do: treat the group as a relationship in its own right, attribute every message in it to the right human, and keep that human as one record when they change handle, DM you privately, or move the conversation to WhatsApp. Pantheon does exactly that — the project, the founders, and the thread stay attached to each other, and the founder with the new handle lands on the same record.

OTC is the most chat-native corner of the industry. Historically a block trade was arranged by pinging several desks in Telegram for a price and taking the best one, and although institutional desks are moving execution onto RFQ platforms, the relationship side has not moved anywhere. The counterparty intro, the "are you a buyer or seller of X this week" check-in, the settlement coordination when a wire lands late — that is still a DM.

What breaks: an OTC desk's most valuable asset is its counterparty list, and that list is a set of Telegram chats in the head of whoever runs BD. When that person is on a plane, nobody else knows that the fund in Zurich only trades on Tuesdays or that the family office wants a call before anything over a certain size. There is also a compliance dimension the desk cannot ignore: when someone asks who you dealt with and what was said, "it's in my Telegram somewhere" is not an answer.

What a Telegram-first CRM has to do: give every counterparty a durable record with the full conversation history behind it — searchable, not screenshotted — and keep that record when the same person reaches you from a second account or on a different messenger. Pantheon keeps the counterparty, the threads, and your notes on one profile, so the desk's memory stops living in one person's phone.

Allocators are arriving. BNY Wealth's 2025 single-family-office study found 74% had invested in or were exploring digital assets, and the path in almost never starts with a data room. It starts with an intro chain: an advisor introduces a principal to a fund manager in a three-person Telegram group, the principal asks two questions, and then nothing happens for four months.

What breaks: the cycle is long and the volume is low, which is exactly the shape a spreadsheet handles worst. The fund manager who raised from one family office last year cannot remember, when the principal resurfaces, whether the concern was custody or liquidity — and the principal, who talks to thirty managers, cannot remember which one had the Asia exposure they liked. Both sides lose the thread, and the relationship restarts from zero.

What a Telegram-first CRM has to do: hold the entire history of a slow relationship so that month four picks up where month one left off, across the DM, the group with the advisor, and the WhatsApp message the principal sent from a different phone. Pantheon keeps that history on the person; when they come back, the context is already open.

For a fund, Telegram is the top of the funnel. A founder DMs after a conference. A co-investor drops a deck into a syndicate group. A scout forwards a message with "worth a look." The data-room link, the round size, and the lead's name all live in a thread — and the thread is one of a few hundred.

What breaks: three months later the partner asks "who sent us this?" and the associate spends an afternoon searching Telegram by keyword. The founder who was passed on in March raises again in October, from a new username, and the fund takes the first call all over again because nothing joined the two conversations. Deal flow that was never written down is not deal flow — it is a feeling that you saw something once.

What a Telegram-first CRM has to do: turn every founder, co-investor, and scout into a record the moment they message, attach the threads and the links, and recognize them when they return under a new handle or on a new channel. Pantheon builds that record from the conversations you are already having, and you can bring your own AI over it to ask what a founder said last time or which deals went quiet.

Ecosystem teams, launchpads, and partnerships leads live in the largest groups in crypto. A partnership starts as a reply in a 200-person ecosystem chat, moves to a five-person working group, and is negotiated in DMs with two of those five. Telegram supergroups can hold 200,000 members, and the useful signal in any of them is a handful of humans.

What breaks: "twenty people, no idea who's who" scales badly. The BD lead knows the three people who matter in each group, but that knowledge is not written anywhere, and when they leave the team the groups stay and the map goes with them. Meanwhile the same integration partner is in four of your groups under one handle and in your DMs under a display name, and your team is treating them as several different people.

What a Telegram-first CRM has to do: figure out who actually said what in a big group, collapse the same human across groups and DMs into one record, and let you find "the person from the wallet team who offered the integration" without remembering their handle. Pantheon attributes every group message to a real contact and keeps the group itself as a relationship, so the map survives the person who drew it.

Crypto capital is cross-border by default. A Dubai-based desk, a Singapore fund, a founder in Lisbon, and a family office in São Paulo do not share an email culture, a working day, or a phone network — but they all share Telegram, the second most popular messaging app in the world outside China. The relationship starts in a group at a conference in Dubai, continues as a DM at 2 a.m. Singapore time, and closes over a WhatsApp number the founder uses for their local bank.

What breaks: three channels, four time zones, and the same person under a Telegram handle, a WhatsApp number, and an X/Twitter profile that nothing connects. Follow-ups get sent to the wrong app. The introduction that happened in the Dubai group is invisible from the Singapore DM. The record, if there is one, is a note in someone's phone that says "Lisbon founder — the one with the good tokenomics."

What a Telegram-first CRM has to do: follow the person, not the channel. Pantheon resolves the Telegram handle, the WhatsApp number, the iMessage thread, and the X/Twitter profile into one contact with one history, so a relationship that spans continents and apps reads as one conversation.

Every desk on this page has the same worst week of the year: the one after a major conference. You come home with forty new Telegram chats, a dozen new groups you were added to at dinners, and a phone full of people whose display name is a first name and an emoji. Half of them said something worth following up on. You remember about six.

What breaks: the follow-up window is short and the context is thin. The market maker who mentioned a listing in Q1, the family office that wanted "a deck, nothing fancy," the founder whose token you liked — all of them are in your Telegram, none of them are labeled, and by the time you have worked out who is who, the moment has passed. The conference paid for itself only if those forty chats become relationships, and the tool for that today is your memory.

What a Telegram-first CRM has to do: turn each of those forty chats into a contact on the day they message you, attach the group they came from and the thread you had, and let you ask a week later who said what. Pantheon builds those records automatically from the conversations themselves, and you can ask your own AI over them: who mentioned a raise, who asked for a deck, who you owe a reply.

None of the failures above is loud. Nobody notices the deal that did not happen because the second message went unanswered, or the counterparty who went to another desk because they had to re-explain themselves. The cost shows up as a general sense that the network is bigger than the pipeline — you know more people than you are working with, and you cannot say exactly who.

Desks try three fixes, and each has a known failure. The spreadsheet dies the first busy week, because it depends on someone stopping mid-conversation to type. The general CRM never sees Telegram, so it fills up with the email relationships that were never the problem. The Telegram-only tool helps until the relationship moves — and in crypto, where a settlement conversation on WhatsApp or a DM on X/Twitter is routine, it moves.

The fix that holds is the one that does not depend on anyone doing anything: the record builds itself from the conversations, follows the person across channels, and is there when you need it. That is what Pantheon is for.

Connect Telegram and Pantheon quietly builds a contact for everyone you actually talk to — already filled in from your history, not a blank form begging for data entry. One inbox across Telegram, WhatsApp, iMessage, and X/Twitter. A profile for each person that keeps itself current and collapses their handles, numbers, and profiles into one record. Groups where we figure out who actually said what. And Milo, your assistant, can recap a noisy thread and remind you who you owe a reply — over your own chats, with the AI you choose. No spreadsheets, no Zapier spaghetti, no "I'll log it later" (you won't).

What it is not: a bot you add to the group, a browser extension, or a trading tool. Pantheon holds no keys and touches no wallets. It is the memory underneath the conversations where your deals already happen.

Those were built for sales teams who live in email and love a tidy pipeline stage. They can't see your Telegram, they expect you to keep everything updated by hand, and they have definitely never survived a 200-person alpha group chat. Pantheon works the other way around: it starts from your messages, fills itself in, and is built for how crypto deals actually go down. Head-to-head: Pantheon vs Salesforce, Pantheon vs HubSpot.

The same failure — nothing is writing it down — shows up differently for each desk. The grid is the short version of the sections above.

Yep. Group chats are fully captured and searchable, and we figure out who actually said what — so "twenty people, no idea who's who" turns into real contacts, and the group itself stays visible as a relationship. Telegram supergroups can run to 200,000 members; the useful signal is a handful of humans, and that is what ends up on your records.

Yes — Pantheon is Telegram-first, so you connect your own account and the record builds from the conversations already on it. No bot in your groups, no extension in your browser. Pantheon is in alpha; connect WhatsApp, iMessage, and X/Twitter the same way.

Your data is yours. It's tied to your account, it is never sold, and Pantheon holds no keys and touches no wallets — this is a memory over your conversations, not a trading tool. The specifics are in our privacy policy.

Pantheon is in alpha. Join the Waitlist and we'll sort out pricing that fits your desk when you're in. The cost worth comparing against is not a subscription — it is the deal you did not follow up on, the counterparty who re-explained themselves to a competitor, and the hour after every conference spent working out who was who. Any tool that removes those pays for itself in a week; the question is whether it does so without asking you to do the data entry.

Today everything on this page works for a single operator with their own accounts. Shared team workspaces are coming soon and are provisioned through guided onboarding rather than self-serve — reach out at enterprise@pantheon.run if you want your desk in early.

It gives you the conversation history behind every counterparty, searchable and on one profile, rather than scattered across chats and screenshots. That is a relationship record, not a regulatory system — your compliance obligations and the tools that satisfy them are yours to define, but "it's in my Telegram somewhere" stops being the answer.

One. Pantheon resolves the Telegram handle, the WhatsApp number, the iMessage thread, and the X/Twitter profile into one contact with one history. When a founder DMs from a new username or a counterparty switches to WhatsApp for settlement, it lands on the same record.

No. Pantheon is the CRM and memory underneath the conversations where your deals happen. It holds no keys and reads no wallets. If that changes it will be said plainly here; for now, the job is remembering the people and what was said.