AI customer service chatbot for email and chat, built on your own knowledge base
Most incoming email and chat inquiries are repeat questions. The customer service AI chatbot and email assistant answer them from your own knowledge base with source citations, and drafts the harder cases for your team. Pilot in 2-3 weeks, implementation in 6-12 weeks; we measure results from day one. Delivered by a Budapest-based team.
The next step is the AI opportunity assessment: from HUF 500,000 + VAT.
Updated:
The customer service AI chatbot answers incoming email and chat questions from your own knowledge base: policies, product documentation, past replies. Every answer arrives with its source, so both your team and your customers can see what it is based on. Anything the system cannot answer reliably is handed to a colleague with a summary and a draft reply. The internal AI knowledge base for your own staff, built on the full document base, is a separate service.
Across our 46+ completed projects, automated processes reach a 40-45% efficiency gain; typical ROI is 6-12 months. At an insurance company, 78% of staff used the internal AI knowledge base built on 12,000 documents within the first month — every answer links back to the original document.
A full team carries the work: a project manager, senior developers, and an AI DevOps engineer. We operate production-grade AI systems, and we run the EU AI Act risk classification at the start of every project.
What the customer service AI takes over
Frequent questions answered automatically
The system answers from your own knowledge base — policies, product documentation, past replies — by email and chat, around the clock. Every answer arrives with its source.
Harder cases drafted for your team
Anything the system cannot close reliably goes to a colleague: it summarizes the case, pulls the related documents, and writes a draft reply. Your colleague reviews and sends instead of writing from scratch.
Source citation on every answer
Each answer links back to the original document — a policy, a product spec, a past reply. The answer can be verified, and outdated content surfaces quickly.
Measured from day one
The pilot sets the baseline: response time, the share of inquiries closed automatically, the number of escalations. After implementation we track the same numbers, so the change is measurable.
How the implementation works
- 1
AI opportunity assessment — from HUF 500,000 + VAT
We review your incoming inquiry types and your existing knowledge base: policies, product documentation, past replies. The output is a written plan: what the system can close automatically, what it drafts for your team, and what is missing from the knowledge base.
- 2
Pilot
We test the system on your most frequent questions in production conditions. A small proof of concept is ready in 2-3 weeks and sets the baseline: response time, share of inquiries closed automatically.
- 3
Implementation and operations
We connect the assistant to your existing email, chat, and CRM systems — without ripping and replacing. Most projects go from kickoff to production in 6-12 weeks. Source code, models, and data remain yours; monitoring and support are part of how we work.
Why Leventech
Our team has built production systems since 2013
46+ completed projects, 96% client satisfaction. What we build, we also operate and maintain.
EU AI Act and GDPR built into development
Risk classification at project start, data protection in the design. Customer service assistants typically fall into the limited-risk category; we also have experience implementing high-risk systems.
Your data stays with you
On-premise or private cloud deployment when needed, with open-source models. Customer data never leaves your environment.
Common questions
An AI assistant or integration project is usually €15-40K. The entry point is the AI opportunity assessment: from HUF 500,000 + VAT — it shows what the system could close automatically in your case before you spend on development. Typical ROI is 6-12 months.
The ones with a documented answer in the knowledge base — policies, product documentation, past replies. Every answer arrives with its source. Where no reliable source exists, the system hands the case to your team.
No. It handles the frequent questions and drafts the harder cases for your team: a summary, the related documents, a draft reply. Your team's time is freed up for high-value interactions.
The pilot sets the baseline: response time, the share of inquiries closed automatically, the number of escalations. After implementation we track the same numbers, so the change is measurable.
When needed, we deploy on-premise or on your private cloud, with open-source models running on your servers. Your data never leaves your environment. Data protection is part of the design, and we run the EU AI Act classification at project start.
A small pilot is ready in 2-3 weeks. Most projects go from kickoff to production in 6-12 weeks.
Still have questions?
Related resources
Start with a conversation
We look at your incoming inquiry types and tell you what the system could handle on its own.
What would you automate in your business?
Tell us briefly about the task. We will reply by email to arrange an initial conversation.
The next step is the AI opportunity assessment: from HUF 500,000 + VAT.
What happens after you send?
In the first conversation we review the task, your existing systems and the outcome you need. Then we discuss whether a detailed assessment would help. We provide a proposal before any paid work.
Or reach us directly:
[email protected]
Monday-Friday, 9:00-17:00 CET