Customer service
AI reads customer messages from multiple channels, detects the topic, and drafts a reply in the company's tone - an employee just approves it. Response time dropped by about 70%.
Companies with high customer inquiry volume
The support team could not keep up: messages from e-mail, forms, and messengers landed in different places, and most were about repetitive matters.
Merging channels into a single queue, message classification, and draft reply generation with human approval.
n8n pulls messages from every channel into one queue. The model detects topic, urgency, and customer data, then drafts a reply in the company's tone with case context. Nothing goes out automatically - an employee approves or edits the draft in one click, and the full history is stored in the database.
Response time about 70% shorter, the team focuses on harder cases, and customers get an answer the same day.
LegalTech / law firms
A self-hosted law firm system: cases, documents with local OCR, paper archive, procedural deadlines, debt collection, and an AI assistant - without sending case files to the cloud.
~90k lines of code
Environmental services
An AI chatbot answering from the company knowledge base, plus a registry of issued and returned equipment. Customer support workload dropped by about 40%.
~40% less support workload
Phone sales and support
Every customer call is automatically transcribed and summarised - with decisions and action items - and the note lands in the CRM. Full contact history without manual notes.
