Sean Assaf's Telebiz runs AI locally in the browser, routing requests directly to model providers, storing only encrypted business metadata. The shift challenges traditional SaaS dashboards and targets the broken workflow of manual CRM entry from messaging apps.
Telebiz, a CRM tool built directly inside Telegram, is pushing dealmakers to stop jumping between chat apps and external dashboards. The company runs AI locally in the browser, connecting straight to Telegram's servers and routing context requests directly to model providers. Nothing goes through Telebiz's own servers, founder Sean Assaf said.
The workflow it targets is familiar to anyone in Web3 business development. A venture partner negotiates a token swap in Telegram. An ecosystem lead hammers out a grant agreement in the same thread. Then both open a separate browser tab, log into HubSpot or Salesforce, and retype what happened. Assaf calls that a broken paradigm.
"Most tools sit between you and your messages – they ingest your chats onto their servers, process them, and hand back an output," he said. "Once you accept that shape, you are choosing between usefulness and privacy, because the vendor has to hold your content to be useful."
Telebiz avoids that tradeoff by storing only the business layer: encrypted chat IDs, the mapping between a chat and a HubSpot deal, follow-up rules, and timestamps. The encryption uses per-organization keys. "The AI gets deep context, and we get none of it," Assaf said.
He expects traditional B2B dashboards to survive only as places to look at reports, not as places to do work. "Manual CRM entry was only ever a workaround for software's historical inability to understand human dialogue," he said. "Once the system can read the exchange and act on it, the data entry layer has no reason to exist."
The shift matters for Web3 partnerships because the space runs on Telegram and suffers from mass automated outreach. Assaf said generic blasts fail not because a machine generated them, but because the sender lacked a real reason to reach out. "The recipient list is small – maybe a few hundred people who genuinely matter for any given partnership thesis, and most of them already know each other," he said. "Blasting that group does not scale your outreach; it burns it, and you only get to do it once."
His operational framework is simple: automate everything that happens before and after a decision, never the decision itself. Retrieval, thread summaries, drafting options, logging agreed terms, and chasing cold follow-ups are administrative tasks. The actual decision to reach out, negotiate, or close stays human. "Teams automate the outreach, which is sensitive, and hand-update the CRM, which is not," Assaf said. "Reverse it, and both halves get better."
He sees a larger gap in Telegram's ecosystem. The platform is built for individuals, but business is a team sport. There is no shared context, no ownership of a relationship, and no clean handoff when someone leaves. "That whole layer is missing, and almost nobody is building it," Assaf said.
If Telebiz succeeds, enterprise software may stop looking like a complex dashboard and start looking like a typical inbox.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.