An LLM agent that qualifies B2B leads from AmoCRM: notes from a production build

Description and use

An LLM agent replaced manual first‑touch handling in AmoCRM, cutting qualification time from 40 to 8 minutes, ensuring every lead receives a reply, and boosting the share of leads progressing to sales conversations by 31%.

Problem

A B2B client had 50‑80 inbound leads per day and a single first‑touch sales manager who spent six hours daily answering repetitive questions, leading to uneven filtering and missed replies.

Outcome

Lead qualification time reduced from 40 to 8 minutes, every inbound lead answered, and the share of leads reaching a sales conversation increased by 31%.

Workflow

• Pull new leads from AmoCRM API v4 each morning
• Classify request type (price question, materials request, etc.)
• Open conversation in the originating channel (Telegram or WhatsApp via Wappi)
• Qualify lead on six criteria (volume, budget, timing, ICP fit, etc.)
• Move deal to appropriate pipeline stage and tag it
• Provide admin panel for managers to view and override conversations

Evidence

A B2B client came to us with 50 to 80 inbound leads a day and one first-touch sales manager. About six hours of that person's day went to the same questions: what are your prices, can you send the deck, do you work with our volume.

Source

Qualifying a lead takes about 8 minutes instead of 40.

Source

The share of leads that reach a sales conversation went up 31%. That came from consistency alone: the same questions, asked every time, with nothing forgotten.

Source

Start from the CRM fields, not the prompt. Decide what a qualified lead looks like as data, then make the model fill exactly those fields.

Source

Give people a visible override. The admin panel was one of the cheapest parts of the project, and it is the reason the team accepted the bot.

Source