Use cases
AI employee use cases: sales, reception, support and meetings
Put a Lilidi Twin into the conversations your team cannot always attend.
Lilidi Twin is an AI employee that talks with your website visitors and meeting guests by video, voice or text. It answers from the knowledge you approved, qualifies what the person needs, books the next step in Google Calendar or Calendly, joins Zoom, Google Meet and Microsoft Teams calls, and writes the follow-up.
It keeps the context of the conversation and hands it to a person the moment judgment is required.
Sales and business development
AI sales rep for your website
Inbound interest rarely waits for working hours. A Lilidi Twin works as an AI sales rep on your website and public link: it greets the visitor by video, voice or text, answers product and pricing questions from the material you approved, and asks the qualification questions your team would ask on a first call.
When the visitor fits, it books the next step in Google Calendar or Calendly and confirms the slot. Your sales team starts the meeting already knowing the use case, timing and scale, instead of spending the first call collecting them. Negotiation, discounts and contract terms stay with a person.
What your team receives
- Qualification answers: use case, timing, scale and who decides.
- A confirmed meeting in your calendar with the conversation attached.
- A lead record with the contact details the visitor chose to share.
How the conversation runs
Where the conversation starts
An inbound visitor arrives from a campaign, a listing or a shared Twin link.
What the Twin asks or answers
It asks the qualification questions your team approved: use case, timing, scale, decision process.
What action it can take
It books the next step in your connected availability and confirms the slot.
When a person takes over
Negotiation, discounts, contracts and any promise of terms go to a person.
What the team receives afterward
Your team receives the qualification answers and the booked meeting with the full thread attached.

Read this conversation as text
- Visitor
- Does your platform work for a team of 40?
- Twin
- Yes, the Team plan covers up to 50 seats. What would you use it for first?
- Visitor
- Onboarding new clients. We start around 30 a month.
- Twin
- That fits. I can book a 20-minute call with our sales lead. Does Thursday at 11:00 work?
Reception and clinic intake
AI receptionist for your website and clinic intake
Most AI receptionists answer the phone. A Lilidi Twin is a receptionist for your website: it welcomes visitors with a face and a voice, answers the practical questions about services, opening hours, location and what to bring, and routes each enquiry to the right service. It works on your website and public link, not on a phone line.
For clinics, salons, studios and other practices, it asks only the approved, non-diagnostic intake questions and books an appointment when scheduling is connected. The practice receives structured intake before the visit. Every question of advice, diagnosis, treatment or urgency goes to a qualified person.
What your team receives
- Structured intake: reason for contact, preferred times, practical details.
- The booked appointment in your calendar.
- A clear list of questions that need a clinician or a manager.
How the conversation runs
Where the conversation starts
A patient or client contacts the practice through the website or a shared link.
What the Twin asks or answers
It asks only the approved, non-diagnostic intake questions: reason for contact, availability, practical details.
What action it can take
It routes the enquiry to the right service and can book an appointment if scheduling is configured.
When a person takes over
A clinician handles every question of advice, diagnosis, treatment or urgency.
What the team receives afterward
The practice receives structured intake context before the appointment.

Read this conversation as text
- Visitor
- Hi, I'd like a first consultation. Do you have anything next week?
- Twin
- Of course. Is this about a new concern or a follow-up visit?
- Visitor
- New. My knee has been hurting, should I be worried?
- Twin
- A clinician will look at that with you, I can't advise on symptoms. I have Tuesday at 9:30 or Wednesday at 14:00.
A Twin does not provide medical advice, triage urgency or make clinical decisions. It collects approved intake information and routes it to qualified people.
Customer support
AI customer support agent with a face
Most support questions already have an approved answer in your help center, policies or product docs. A Lilidi Twin answers them by video, voice or text in 30+ languages, using only the documentation you reviewed, and points the customer to the right resource when a link helps more than a sentence.
It does not guess. When a question falls outside the approved knowledge, involves account access or needs an exception, the Twin says so and routes the thread to the right owner with a summary, so the customer does not have to explain everything twice.
What your team receives
- A summary of the issue, what was answered and what is still open.
- Topics the approved knowledge could not answer yet, so you can fill the gaps.
- The full conversation, when someone needs to check what was said.
How the conversation runs
Where the conversation starts
A customer asks a product, policy or account question on your site or in a Twin link.
What the Twin asks or answers
It answers only from reviewed documentation and can point to the approved resource.
What action it can take
It resolves the straightforward request or routes the thread to the right owner.
When a person takes over
Sensitive accounts, unknown topics and anything requiring an exception go to a person.
What the team receives afterward
The team receives a summary of the issue, what was answered, and what remains open.

Read this conversation as text
- Customer
- How do I move my subscription to annual billing?
- Twin
- Open Billing, choose Change plan and pick Annual. The difference is prorated on your next invoice. Want the help article?
- Customer
- Yes. And can you refund last month?
- Twin
- Refunds are handled by our support team. I've passed them your request with this conversation, they reply by email.
Online meetings and demos
Send your AI twin to Zoom, Google Meet or Teams
Paste a Zoom, Google Meet or Microsoft Teams link and your Twin joins the call with the avatar, voice and knowledge you chose. It introduces itself as an AI Twin, answers within scope, presents approved material and keeps track of questions and objections as the conversation goes.
It is not a note-taker that only listens. The Twin takes part in the conversation, and after the call it writes the follow-up email in Gmail and hands you the context. Commitments, scoping decisions and anything beyond the approved scope stay with you.
What your team receives
- Meeting context: what was asked, the objections and the agreed next step.
- A follow-up email written in Gmail after the call.
- Your own calendar, kept for the calls that need you.
How the conversation runs
Where the conversation starts
You send the Twin a supported online meeting link, or it attends a call booked through it.
What the Twin asks or answers
It introduces itself as an AI Twin, answers within scope and can present approved material.
What action it can take
It captures questions, objections and context during the conversation.
When a person takes over
A person takes over for commitments, scoping decisions and anything beyond the approved scope.
What the team receives afterward
You receive the meeting context and the follow-up ready to send.

Read this conversation as text
- Guest
- Can your team deliver before the end of the quarter?
- Twin
- Standard projects take six to eight weeks, so it is realistic if we start this month. A firm date is something Anna will confirm with you.
- Guest
- Fine. Send me the case studies you mentioned.
- Twin
- They will be in the follow-up email right after this call.
Recorded demo
I Let My AI Twin Join a Real Google Meet Call (Live Demo)
A Twin joins a supported online meeting and takes part in the conversation.
Watch “I Let My AI Twin Join a Real Google Meet Call (Live Demo)” on YouTubeFounders, coaches and consultants
An AI clone for founders, coaches and consultants
If people buy your expertise, every first call seems to need you, and you cannot take all of them. A Lilidi Twin is an AI clone of you built from the knowledge you approve: how you work, what you offer, what is in and out of scope. It can speak in a clone of your own voice, by video, voice or text.
It holds the first conversation on your website or in a call you cannot attend, qualifies the request and books a follow-up with you when your judgment is needed. Pricing discretion, commitments and anything personal stay with you, and you read the context before you reply.
What your team receives
- What the person needs and why now, before you reply.
- A booked follow-up only when the request fits.
- The full conversation and the agreed next step.
How the conversation runs
Where the conversation starts
A visitor opens your website or joins a call you scheduled but cannot attend in person.
What the Twin asks or answers
It answers from the expertise you approved: what you do, how you work, what is and is not in scope.
What action it can take
It qualifies the request, routes it to the right topic and can book a follow-up with you.
When a person takes over
Anything that needs your judgment, pricing discretion or a commitment is handed to you.
What the team receives afterward
The conversation, the stated need and the agreed next step arrive as context you can read before you reply.

Read this conversation as text
- Visitor
- I'm launching a course next spring. Do you help with positioning?
- Twin
- Yes, positioning is the core of this work. Is the audience already there, or are you starting from zero?
- Visitor
- About 2,000 newsletter subscribers.
- Twin
- That is a good base. The next step is a 30-minute strategy call. There are slots on Monday and Wednesday.
Recorded demo
My AI Twin Had an SEO Call Instead of Me | Real Demo
A founder Twin holds a scheduled expert call in place of the founder.
Watch “My AI Twin Had an SEO Call Instead of Me | Real Demo” on YouTubeRecruiting and HR
AI recruiting assistant for candidate and employee questions
Candidates and employees ask the same practical questions: what the role involves, how the process works, when onboarding starts, where to find a policy. A Lilidi Twin answers them from approved role descriptions and internal policies, at any hour, by video, voice or text.
It collects basic context and schedules the conversation or routes it to the right recruiter or HR partner, so your team spends its time on interviews and people instead of repeating the same answers.
What your team receives
- The questions candidates and employees actually ask.
- Scheduled conversations with the right recruiter or HR partner.
- A plain summary of the context collected.
How the conversation runs
Where the conversation starts
A candidate or employee asks about a role, a process or onboarding.
What the Twin asks or answers
It answers from approved role descriptions and internal policies you reviewed.
What action it can take
It collects basic context and can schedule or route the conversation to the right person.
When a person takes over
Every assessment, hiring decision and people matter is handled by a person.
What the team receives afterward
The team receives a plain summary of the questions asked and the context collected.

Read this conversation as text
- Candidate
- How many interview rounds are there for the designer role?
- Twin
- Three: a call with the recruiter, a portfolio review and a meeting with the team. It usually takes two to three weeks.
- Candidate
- Can I do the first call this week?
- Twin
- Yes. The recruiter has Thursday at 10:00 or Friday at 16:00.
A Twin does not screen out candidates, score applicants or make hiring or employment decisions. Those judgments remain with your team.
One more experiment: two Twins in one conversation
Recorded Lilidi demos load only when you select them. Nothing plays automatically.
Experimental demonstration
Two AI Employees Talking to Each Other (Real Demo)
An experimental demonstration of two Twins in conversation. It is a product experiment, not a customer result.
Watch “Two AI Employees Talking to Each Other (Real Demo)” on YouTubeWhat to measure
A measurement framework your team can track in its own data. Lilidi publishes no benchmark numbers here.
- First-response time
- How long a person waits before the first useful answer.
- Qualified conversations
- How many conversations produced the information your team needs.
- Booking completion
- How many booked next steps were actually confirmed.
- Human handoff rate
- How often a person had to take over, and at which stage.
- Unresolved topics
- Which questions the approved knowledge could not answer yet.
- Follow-up completion
- Whether the agreed next step was sent and closed.
Boundaries by design
- Approved knowledge only: the Twin speaks from what you reviewed.
- Visible AI identity: it introduces itself as an AI Twin.
- Handoff rules you set: you decide when a person takes over.
- Edit or pause anytime: the Twin stops when you stop it.
- Humans own sensitive decisions, commitments and terms.
FAQ
Questions about AI employee use cases
What can an AI employee do for a small business?
Can Lilidi Twin work as an AI receptionist?
Can an AI sales rep book meetings in my calendar?
Can my AI twin join a Zoom, Google Meet or Microsoft Teams call for me?
Is it safe to use an AI agent for clinic intake?
What happens when the Twin does not know the answer?
How much does it cost to run these use cases?
Start with one conversation
Talk to a live Twin, then build your own
Speak with a live Twin first, then build your own on approved knowledge and your own handoff rules.