Live triage

AI triage for repair and service requests

The right answer, the right party, in a single phone call. For housing associations, property maintenance companies and HOA managers.

Book an introduction
Live triage
  1. 01

    The request comes in

    Tenant, address and open job sheets are ready before the call is answered.

  2. 02

    Flixer listens in live

    The call is matched against the knowledge base. The right follow-up questions appear, starting with whether the problem can be solved on the phone.

  3. 03

    The follow-up is ready

    The tenant, your own team or the right partner. The work order or partner email is prepared.

The knowledge exists, just not at the right moment

Whoever answers the phone rarely has the technical knowledge to ask the right questions and route the request.

  1. 01

    Tenant calls

    reports ‘the boiler is not working’

  2. 02

    Call handler

    lacks the technical knowledge, records only the complaint

  3. 03

    Planner

    sends an engineer, without a fault code or context

  4. 04

    Engineer

    arrives at the door with the wrong equipment

18%

of repair requests cannot be fixed in one visit, purely because of what is not asked during the phone call

Nobody can keep all of this in their head

Every housing association has its own policy rules, building agreements and supplier contracts, and nobody can have all of that to hand. The problem is access to that knowledge at the moment of the call, not the employee who answers.

A technical colleague listening in

Your employee runs the call, Flixer supplies the knowledge. Three moments from the same call.

Before answeringContext ready

The request comes in

Tenant, address and open job sheets are ready before the call is answered.

Book an introduction
Screen showing an incoming call from a known tenant, with name, address and a summary prepared by the assistant.
During the callLive coaching

Flixer listens in live

The call is matched against the knowledge base. The right follow-up questions appear, starting with whether the problem can be solved on the phone.

Book an introduction
Screen with the live transcript of the call on the left and live coaching on the right: topic detected as a heating pressure issue, with the follow-up question about the pressure gauge.
After the callFollow-up

The follow-up is ready

The tenant, your own team or the right partner. The work order or partner email is prepared.

Book an introduction
Screen showing the follow-up action after the call: a recommended email to the partner, with the prepared email already drafted.

The knowledge base is the core

Internal policy and trade knowledge, translated into questions and decision rules the system applies live. The more data, the better it gets.

Diagram of the Flixer knowledge base. On the left, what is provided at the start: demarcation and decision rules, the organisation's expertise and system integration. On the right, what it learns from practice: all service requests, work orders and asset data. Below, shared knowledge: anonymised repair and maintenance insights.

Provided at the start

  • Demarcation & decision ruleswho pays and who acts for an HVAC failure, a leak, a lift malfunction
  • The organisation's expertisethe experience of the service desk and technical managers, captured
  • System integrationdata from your property management system flows in automatically via the API

Learns from practice

  • All service requestsevery request feeds the system
  • Work orderswhat was actually done on site
  • Asset datawhich installation, supplier and contract belong to which building

Shared knowledge: anonymised repair and maintenance insights, benefiting every customer, never client-specific data.

What the missing knowledge costs

A model at 10,000 requests a year. During shadow mode we measure your own figures.

Your assumptions

Adjust the three values to match your organisation.

Total number of requests on an annual basis.
Share of requests that leads to extra work.
Average cost of an extra visit, in euros.
Avoidable costs per year € 270.000 per year

Requests per year × percentage not resolved in one visit × cost per unnecessary visit

1.800 unnecessary visits per year

This is a model based on our own field figures, not a promise. During shadow mode we measure your own figures.

Where the money leaks away

Wrong engineer

Drives to a job they cannot finish. The request goes back into planning.

Warranty missed

Your own engineer on work the supplier should have paid for.

Repeat callers

Whatever is not asked properly the first time calls back within a week.

The programme in three phases

Each phase has its own go/no-go, and minimal work for your IT department.

  1. Shadow mode

    Six weeks

    Listening in on the repair line. You receive a report on your own requests: where it goes wrong and what can be saved. Measured, not estimated.

    What IT has to do

    One connection to the existing phone line.

  2. Implementation

    Four weeks

    Your maintenance policy becomes a working knowledge base, with live triage on the repair line. Followed by two weeks of tuning.

    What IT has to do

    Nothing new. An ERP connection needs one-off API access and can follow later.

  3. Pilot

    Six months

    Live on every request, measured against targets agreed before the start and against the baseline from shadow mode.

    What IT has to do

    Nothing.

Security, GDPR and call recordings

What is settled before a single request goes through the system. The full policy is shared under NDA.

Data processing agreement

You sign a data processing agreement with Flixer Pro as processor and your organisation as controller. Sub-processors are named in its annex.

Where the data sits

Application, database and storage run inside the EU (Ireland). Which suppliers those are, and the region each processes in, is listed per sub-processor in the annex to the data processing agreement.

Encryption and separation

TLS 1.2 or higher in transit, AES-256 at rest. Customer data is separated at the database layer with row-level security, not only in the application.

Access and logging

Role-based permissions, two-factor authentication you can enforce across your organisation, and single sign-on. Login and change logging, with an exportable overview of users and permissions.

Incidents and breaches

A documented incident response policy with defined severity levels. Notification to the Dutch DPA within 72 hours where required, and you hear from us without undue delay if your data is affected.

Policy and review

The security programme is built to the requirements of the GDPR (Article 32), SOC 2 and ISO 27001. The full policy, including what is still in development, is shared under NDA.

Full information security policy, data processing agreement and sub-processor list available on request.

Frequently asked questions

Does this replace our call handlers?

No. Your employee runs the call; Flixer supplies the knowledge at the moment it is needed. The follow-up questions and the suggested next action appear during the call, and the employee decides.

Is there a data processing agreement?

Yes. You sign a data processing agreement in which Flixer Pro is the processor and your organisation the controller. Sub-processors are named in its annex. The full information security policy is shared under NDA.

Where is our data stored?

The application, database and storage run inside the EU (Ireland). The full sub-processor list and the region each processes in is in the annex to the data processing agreement, so you can review them by name before signing.

Are you ISO 27001 certified?

Our security programme is built to the requirements of the GDPR (Article 32), SOC 2 and ISO 27001, and the underlying hosting infrastructure is ISO 27001 certified. Flixer Pro itself does not currently hold its own ISO 27001 or SOC 2 certificate.

Does it work alongside our existing ERP?

Yes. Flixer works alongside your existing software. You can start without an integration: during shadow mode, one connection to the existing phone line is enough. An ERP integration needs one-off API access and can follow later.

Sound familiar?

See live triage on your own requests.