Not another tool.
An employee that does the work.
lixwi agents take a whole function off your team: the request comes in, the agent triages it, opens the record, coordinates the people or suppliers involved and follows it through, inside the systems your company already runs.
Built to amplify your people. Your team keeps the judgment, the agent carries the volume.
Live in 15 business days · Managed or self-hosted · Every action logged and approvable
Works inside the systems you already run
You are not buying software. You are buying
a function that runs itself.
Companies where the work is already defined and already staffed.
Industry matters less than the shape of the work. We look for a function with clear rules, real volume and people already doing it, then build the agent around how your company runs it rather than a generic template.
- A function that already runs on written rules and repeats every day
- Two or more people spending most of their week on coordination
- Work that lives across a few connected systems, not one clean database
- Requests that arrive after hours and wait until morning
- Growth that currently means hiring another coordinator
A chatbot takes a message. The agent closes the loop.
Pick a situation your team handles every week and see what changes.
Takes a message and leaves it in an inbox. Someone reads it at 8am and starts from scratch.
Acknowledges the customer, classifies urgency against your rules, opens the job in your system of record, checks supplier availability and confirms a morning window before your team logs in.
You decide exactly what it is allowed to do.
Every action sits in one of three buckets, set during onboarding and changeable at any time. Nothing moves between them on its own.
- Acknowledge an incoming request the moment it arrives
- Classify and prioritise it against your own rules
- Open the ticket, job or record in your system of record
- Coordinate with suppliers or teams below your spend threshold
- Confirm timing and next steps with the customer
- Follow up, chase and keep every record current
- Anything above the spend threshold you set
- Engaging a supplier outside your approved list
- Anything with legal, safety or compliance exposure
- Reopening or reversing work already signed off
- Any commitment made on a client's behalf
- Approving or issuing payment of any kind
- Interpreting or amending contract terms
- Making legal or regulatory determinations
- Handling safety-critical events on its own
- Speaking for the company to an owner or investor
Every action is logged with its trigger, the authority it was taken under, the model used and the human approver where one applied. Logs are yours to export.
The agent remembers how your company actually operates.
Most AI starts every interaction from nothing. Contextor is the persistent operational memory underneath every lixwi agent: organisational memory, not conversation history. It is why the agent behaves like it has held the role for a year rather than five minutes.
The longer we run your account, the better the judgment gets.
- That this supplier missed the last three appointments in that region
- That this customer flags everything as urgent, and what actually happened each time
- That this account has a standing instruction outside your default rules
- That this issue has recurred three times and needs a different fix
What changes in the first quarter
We measure the same things you already report on: requests handled without a human, time from request to dispatch, backlog, and after-hours resolution.
Coordination hours back
Intake, triage, dispatch and follow-up stop consuming full-time roles.
After-hours covered
Overnight and weekend requests are resolved before the next business day, not queued as messages.
Volume without headcount
Take on more customers, sites or accounts without adding a coordinator for every increment.
Priced against the work, not per seat.
You pay for a function being performed, metered on the volume that function actually carries: units, sites, tickets or accounts, whichever your company already counts.
Single agent
From $900 per month
Volume metered · $3.500 onboarding · e.g. $0,55 per unit for property management
Start with one agent on one function. Every lixwi agent works standing alone, so nothing else has to be in place first.
Multi-agent bundle
From $3.100 per month
Volume metered · $8.000 onboarding · roughly 25 percent below buying separately
Several agents sharing one operational memory, so coordination, support and back-office work stop handing tasks back to a person in between. Expand once the first agent has proved itself.
Annual contracts, billed monthly, with a defined task volume allowance. We quote against your function after a short scoping call.
Stated precisely, not marketed loosely.
Handing an operational function to software is a real decision, so we describe exactly what happens to your data rather than reaching for a slogan.
Self-hosted deployment is available for companies that need everything to stay inside their own infrastructure.
- Your source records stay in your platform, email and document systems. lixwi does not keep a bulk copy of them.
- Access is over authenticated, scoped, revocable API connections, encrypted in transit.
- Record content is retrieved into working memory to perform a task and passed to the selected model provider under that provider's terms.
- Operational memory created by doing the work, such as task state, decisions and notes, is retained for the contract term and deleted within 30 days of request.
- On the self-hosted tier everything above runs inside your own infrastructure and lixwi retains no customer data.
One agent now. A workforce over time.
Contextor, the event substrate and the orchestrator underneath are industry neutral, so an agent proven in one operation can be shaped to another. Each one earns its place by performing, in sequence, rather than being announced early.
An operational expansion tool, not a headcount-reduction product.
- Proven in production
Coordination Agent
Request intake, triage, job creation, supplier dispatch, customer communication and escalation. First deployed in residential property management, where it runs maintenance end to end.
- Available now
Support & Communications Agent
Inbound message handling, status responses, routing and follow-up across email, chat and phone, working from the same operational memory.
- Next
Back-office Operations Agent
Bounded finance and admin work: reconciliation prep, invoice chasing and document handling, scoped around what still requires a qualified person.
- Later
Custom agents for your function
Where a function is well defined and already staffed, we build the agent for it on the same platform rather than asking you to change how you work.
The short version.
What is an AI employee?
A software agent that performs a defined operational function inside your business instead of assisting the person who performs it. You buy the work completed, not a tool your team has to learn and drive. It is not a chatbot and it is not a copilot.
Does this only work for one industry?
No. The platform is industry neutral: any function that is high volume, rule driven and already staffed is a candidate. Our first production deployment is maintenance coordination for residential property managers, because it was the fastest place to prove an agent can hold a job to a measurable standard. The same agent pattern runs in operations, support, logistics and back office.
Does it replace my team?
It takes the repetitive intake, triage, dispatch and chase work off them. Judgment calls, relationships, escalations and anything with legal or safety exposure stay with your people. Companies typically redeploy that time into growth rather than cutting roles.
What happens to my data?
Your source records stay in your systems. lixwi does not maintain a bulk copy of them. Record content passes through lixwi and the selected model provider while a task is being performed, and operational memory created by doing the work is retained for the contract term. A self-hosted tier keeps all of it inside your own infrastructure.
How long does onboarding take?
Fifteen business days from contract to supervised operation is the target. Your side is roughly 8 to 12 hours: a rules and thresholds workshop, validation of your supplier and contact lists, system access, and a review session at the end of the supervised period.
See the agent run on your operation
Tell us which function eats your team's week and which systems it runs on. We will walk you through the agent handling real requests, with your own rules and spend thresholds applied.