Everyone gets their own agent
It lives in the phone, the messenger and the operating system. It knows its owner’s calendar, budget and habits — delegating is simply faster than searching.
Roadmap
Soon your client will be brought in by their agent
Everyone is about to have a personal AI agent: it will book the appointment, reserve the table and buy the tickets on its owner’s behalf. We are making sure your business is the one that agent can reach — support for the MCP and A2A protocols is already in the works.
How it will work
In development The client’s agent
Client
Book me a haircut with Anna, Thursday evening
Their agent
Booked: Thursday 7:30 pm — haircut with Anna, $32
Your business on LALAPAM
- Services and prices list_services
- Open slots find_slots
- Rules and discounts apply_rules
- Payment create_payment
What is about to happen to booking
Ordering stops being a trip to a website
A person says “book me with my usual stylist on Friday evening” and goes back to work. No search, no slot picking, no form, no manual payment.
Demand flows to whoever the agent can work with
An agent will not parse your page layout and will not call your front desk. It picks the business whose services, slots and prices it can read and book over a protocol.
Four things it gives a business
A demand channel that is still empty
While the market argues about whether it needs AI, you end up among the few an agent can approach directly.
Booking without friction
An agent never gets distracted, never finds the form awkward on mobile and never puts it off. A started conversation ends in an order.
Your rules stay yours
The agent sees the same prices, slots and limits you set in the admin panel. No middleman rewrites your price or resells your own clients back to you.
Being invisible costs more than a subscription
A business without a protocol will look like a company without a website in 2005: it exists, but nobody finds it.
Where we are now
Already working
- The AI agent books clients in chat, around the clock
- It remembers regulars: their usual service, specialist and location
- It hands the conversation over when a human is needed
- It takes payment inside the conversation
- It acts on your rules, prices and schedule
In development
- An MCP server for your business Catalogue, open slots, prices and order creation — exposed as tools an outside agent can actually use.
- A2A between agents The client’s agent negotiates directly with your business agent: it asks, yours answers and books.
- A card for discovery A description of your business that lets an agent find you and understand what you do and on what terms.
- Permissions and limits for outside agents What an outside agent may do on its own, what needs confirmation, and what stays with a human.
We put no dates here. An item moves to “Already working” only when you can actually use it.
MCP and A2A in plain words
MCP — a power socket for agents
Model Context Protocol: the shared standard an AI agent uses to plug into someone else’s data and actions. For you it means the client’s agent sees your catalogue, your open slots and your prices — and creates the booking under the same rules a person follows in the app.
A2A — the language of negotiation
Agent-to-Agent: the standard agents use to talk to each other. The client’s agent states the task in human terms, your business agent answers with what the schedule really holds and carries it through to a booking.
Why open protocols instead of our own API
A closed integration ties you to one partner and dies with it. An open protocol works like email: connect once and every agent understands you, whoever built it.
The future is easier to meet already connected
Start with what works today: the AI agent books clients in chat right now. The protocols land on top — no migration, no reworking your processes.