AI agent for your business: software that works inside your tools, under your supervision.
Merlin builds AI agents plugged into the tools you already use (CRM, e-mail, spreadsheets, ERP): they read, sort, qualify, prepare, and hand over to your teams for the decision. Nothing goes to production without your approval.
What is an AI agent, and what does it actually do?
An AI agent is software that performs tasks with supervised autonomy inside your tools. Its role: reading an e-mail, qualifying a request, preparing a reply or a quote, recording the result in the right place. Where a chatbot only answers, an agent acts.
It acts, it does not just answer
A chatbot holds a conversation. An agent performs actions inside your tools: creating a record, sending a follow-up, updating a file, routing a request to the right person.
Plugged into your existing tools
CRM, ERP, e-mail, Slack, Notion: the agent settles into your environment. No migration, no new tool for your teams to adopt.
Human supervision built in
The agent prepares, you approve: confidence thresholds, validation queues, an activity log. You decide what goes out without review, and that scope grows as trust builds.
Measured from the first month
Every agent ships with its indicators: volume handled, time recovered, human takeover rate. That is what lets you decide what comes next, with numbers in hand.
What an AI agent can take over for you.
E-mail triage and replies
The agent reads incoming messages, answers the simple requests and passes the rest to the right person, with a summary and a suggested reply.
Lead qualification
Every inbound lead is enriched, scored against your criteria and recorded in the CRM with a summary. Your sales team calls, they no longer type.
Quotes and sales responses
The agent assembles a first draft from your documents and history: a human reviews, adjusts and sends. That is how our tender-response agent works.
First-line customer support
Answers grounded in your knowledge base (RAG), with automatic escalation to your teams as soon as confidence drops or the topic is sensitive.
Document extraction and entry
Invoices, purchase orders, contracts, CVs: the agent extracts the useful data and records it in your tools, with no manual re-entry.
Monitoring and reporting
The agent collects, cross-checks and summarises the information that matters to you (market, competitors, internal indicators) and delivers it at the pace you choose.
How much does an AI agent cost? Our ranges.
The same public ranges as on our general price list, because a provider that hides its prices wastes your time. To understand what these amounts cover, read How much does an AI agent cost?
| 30-min diagnostic | Free |
|---|---|
| Process audit | from €1,500 |
| Custom AI agent | €4,000 – €20,000 |
| Production support | from €300 / month |
Indicative ranges: every project is quoted after the audit. What moves the price: the number of connected tools, volume, how critical the process is, hosting.
Firm quote after the audit: no time-based billing, no surprises.
From audit to a running agent, in four steps.
Audit
We map your processes and cost the 2-3 most profitable cases. The plan is yours, whether you continue with us or not.
Design
Each agent is specified: scope, connected tools, supervision rules, confidence thresholds, error cases.
Build
The agent is developed in your environment and tested on your real data. A demo every week, nothing in production without your approval.
Production
Monitoring, alerts, fixes, evolutions. A monthly review of time recovered and the next cases to automate.
The week-by-week method is detailed on the home page. About six weeks for a first agent.
Your questions about AI agents: role, price, supervision.
How much does an AI agent cost?
A custom AI agent costs between €4,000 and €20,000 depending on scope: number of connected tools, volume, how critical the process is. Production support starts at €300 per month. Our ranges are published in the Pricing section and every project is quoted as a fixed price after the audit.
What is the role of an AI agent?
Its role is to perform tasks inside your tools, with supervised autonomy: reading an incoming e-mail, qualifying a request, preparing a reply or a quote, recording the result in the CRM. It takes the repetitive work, your teams keep the decisions.
What is the difference between an AI agent and a chatbot?
A chatbot answers questions in a conversation. An AI agent acts inside your tools: it creates a record, sends a follow-up, updates a file. Many of our projects combine both: a chatbot up front, agents doing the work behind it.
What is the difference between AI and generative AI?
Generative AI produces content (text, images, code) from a prompt. An AI agent builds on those models but goes further: it chains actions inside your tools to complete a task end to end, with rules and guardrails.
How long does it take to deploy an AI agent?
Our method goes from audit to production in about six weeks for a first agent. A complex agent, or one connected to many tools, can take up to eight weeks.
Can an AI agent run without human supervision?
Technically yes, but we advise against it for processes that commit your business. Every agent we deploy prepares the work and hands over: confidence thresholds, validation queues, an activity log. You decide what goes out without review.