Skip to content
Black Sheep
← Our services

AI agent

An AI agent for your customer inquiries.

Black Sheep builds AI agents that answer using your company’s information. We work with construction companies, manufacturers and new owners, among others, to choose a first use: answering questions, booking appointments or following up on quotes, with a planned handoff to your team.

Discuss your project

YOUR AI AGENT

Answer. Gather. Hand over.

BLACK SHEEP
ILLUSTRATIVE EXAMPLE

An incoming enquiry

The customer asks. The agent prepares the next step.

CUSTOMER

I’d like to discuss an estimate for my project.

AI AGENT

What kind of work are you planning, and in which area?

Your team takes over

The project details and contact information go to the right person to confirm what comes next.

A defined role · a human handoff

The offer

Five agents for different customer needs

01

Website FAQ agent

A chat agent answers using your services, hours and policies. It collects useful details about the inquiry, with a handoff to your team built into the process.

02

Appointment booking agent

The agent offers available time slots, confirms the customer’s choice and sends a reminder. The setup depends on what your calendar supports and the booking rules we agree on together.

03

Text-message agent, on request

Customers text your number. The agent replies using the schedule and instructions provided, then can add useful details to the work order if the connection is possible.

04

Phone-answering agent, on request

The agent takes the call, asks the agreed questions and sends your team a summary. We define permitted topics and how to reach a person based on your availability.

05

Lead follow-up agent

The agent follows up on unanswered quotes according to the agreed schedule and consent requirements. It notifies your team when a customer replies and honours requests to stop.

Our approach

Support from the first step onward.

01

Analysis

We review your business, recurring questions and the tools your team uses. Together, we choose the first agent and priority channel, then define the information needed, consent requirements and rules for handing off to a person.

02

Implementation

We prepare the information, configure the agent and build integrations once compatibility is verified, with only the access needed. Your team tests responses, permitted actions and the handoff to a person before launch.

03

Continuous improvement

We review the conversations selected for follow-up, correct responses and provide support within the agreed scope. When a new model looks relevant, we evaluate it against your use cases and discuss adoption with you before making a change.

A few practical details.

Which agent should we start with?

We start with frequent requests that take up your team’s time. A FAQ agent may fit if the same questions keep coming up; a booking agent may be more useful if scheduling involves many exchanges. We check the available information and tools before choosing the first agent.

When does a person take over the conversation?

The agent can make mistakes. We plan for a handoff when an answer is uncertain, a request falls outside the agreed scope or a customer wants to speak to someone. Sensitive decisions, such as special pricing or a commercial commitment, stay with your team. The handoff process takes your availability into account.

Can we keep our current tools?

We start with your website, calendar and existing customer tracking tools. We check available connections before confirming an integration and limit access to what is needed. If a connection is not possible, we propose another way to pass the information along.

What information can the agent use?

Together, we select useful sources and the information needed. Consent, retention and access are defined before launch. For text messages and follow-ups, we also define who may be contacted and how they can stop receiving messages.

What happens after launch?

The agreed follow-up covers reviews of selected conversations, corrections and ongoing support. Changes to your services or policies may call for an update. We evaluate new models only when they are relevant to your use cases; adoption is decided with you, after testing.

How is pricing set?

Pricing depends on the selected channel, the connections to build, the information to handle and the support requested. After the analysis, our proposal sets out implementation, usage and support costs.

The next step

AI that supports your customers.

Customer inquiries don’t stop when your workday ends.

Discover, at no cost, how an AI agent can answer customers, help with bookings and pass useful information to your team.

Explore for free