Book a call

Support requests

Every request reaches the right team – with full context.

Classify support requests, enrich them with context and route them to the right resolution path.

Book a call

30 min · free · no prep needed

  • Comes in viaEmail, API
  • You decide onUrgent or unclear
  • Lands inService system
Before → After · Support requests

Today: lots of manual work.

  • Requests land in a shared inbox and are distributed by hand.
  • Every ticket is missing customer data, the service agreement or history.
  • Urgent cases get lost in the crowd.
  1. Shared “info@” inboxError reports, orders and ads – all in the same inbox by hand
  2. Read and guessWhich device? Which customer? Somewhere in the third paragraph by hand
  3. Digging out the service agreementCustomer number, maintenance agreement and history from several files by hand
  4. Forwarded by emailNo ticket: the email gets forwarded with “Can you take a look?” by hand
  5. Callback with no contextThe tech team calls and asks everything all over again by hand

Example: Support request SR-2026-1307 from Donau Logistik GmbH – waiting in the inbox

Custom

Describe it, don’t code it.

This is how the XFLOWLY Agent builds the “Support requests” flow – in conversation, tested with a sample document. Your rules, your systems.

Support requestsDraftOverviewSpecial cases
Instructions
The flow takes shape as you describe it.

Example with synthetic data – on a call, we’ll show it with your own document.

Book a call
Your systems

Works with what you already have.

From SAP to the Excel sheet: XFLOWLY connects to your systems and replaces none of them. Highlighted are the ones typically involved in support requests.

S/4HANAECC / R/3Business OneBusiness CentralDynamics 365Navisionabas ERPproALPHAInforOracleOdooweclappXentralWawiSage
DATEVNTCSLexwareRZLDiamantSalesforceHubSpotPipedriveShopwareShopifyMagentoDocuWared.velopELOM-Files
SharePointOneDriveOutlookExchangeTeamsGmailExcelGoogle SheetsServiceNowJiraZendeskFreshdeskPersonioSuccessFactors

And everything without a big name:CAQ systemMESTime trackingPayroll softwareREST & ODataSQL databaseSFTP & network drivePortal without an APIIn-house software

Your system isn’t listed? XFLOWLY still connects – via API, database, file or directly through the user interface. Nothing gets replaced.

Check the connection on a call
Experience built in

Typical challenges – planned for from day one.

We know from experience what tends to go wrong with support requests day to day. The flow catches it.

Content & discrepancies

“Doesn’t work.” – that’s all the email says.

How the flow handles it

The flow asks specifically for serial number, symptoms and photos – tailored to the product.

What the flow covers

  • Take in support requests from shared channels
  • Identify customer, device or serial number, symptoms and urgency
  • Find duplicates and request missing information
  • Assign the ticket to the right team or queue
  • Track SLA response times, resolution and handovers
Practical example · Mechanical engineering

Machine down at a customer: urgent requests reach the right technician in under an hour.

  • Mechanical engineering
  • Packaging machinery manufacturer
  • ≈ 280 employees
  • Service in more than 30 countries
  • ≈ 1,300 service requests per month

  1. 1Starting point

    Service requests came in by email, web form and phone notes, in several languages. Serial number and fault description were often missing; urgent breakdowns sat in the shared inbox next to spare-parts questions.

  2. 2Implementation

    • Requests are read in any language and classified as fault, spare part, maintenance or question.
    • Serial number, machine type and maintenance status come from the ERP; anything missing is asked for specifically.
    • Urgency is assessed from content, customer and service agreement; breakdowns go straight to the on-call team.
    • Tickets are created in Zendesk and assigned to the right team – with solution hints from past cases.

    ConnectedZendeskSAP S/4HANAMicrosoft Teams

  3. 3Outcome

    • ≈ 60%of requests assigned without manual pre-sorting
    • ≈ 100 hrsless manual work per month
    • < 1 hruntil a breakdown reaches the right technician

    Time saved calculated with 8 min of manual work per transaction before.

Project timelineLive after 4 weeks

  1. Week 1WorkshopCategories, urgency rules and teams defined.
  2. Weeks 2–3Test with real inquiriesClassified two weeks of requests in parallel and compared the results.
  3. From week 4LiveLive in all languages.
Good to know

Common questions about support requests.

Rainer GruberAlexander TraxlerReinhard SchnetzingerCan’t find your question?

Ask the founding team directly – on a call or with a quick email.

Book a call
office@xflowly.com

From your systems: CRM data, service agreement and installed base are pulled into the request – including past tickets and matching solutions.

By content, customer and service agreement – for example downtime, safety relevance or agreed response times. Urgent cases get high priority right away and go to the right team.

With topic classification and an assessment of urgency and impact. Contract matching checks the service entitlement, serial number matching the machine or license. Then responsibility-based routing takes over – first level or second level. If the service contract has expired or the serial number is unknown, the team sees it right away.

SLA matching starts the SLA clock for each ticket, and response time and resolution time monitoring run alongside; while waiting for customer response, the clock pauses according to your rules. If an SLA is about to be breached or a fault clearance time is about to expire, the escalation check raises the escalation level. If a ticket lands in the wrong escalation group, it is rerouted. The flow assesses SLA measurement, technical urgency and contractual entitlement separately.

If the same issue comes in across several channels, the flow detects the duplicate ticket and suggests merging the tickets or linking them as parent and follow-up ticket. If a customer follows up on a solved case, the ticket is reopened instead of created anew.

Yes. Knowledge article matching searches your knowledge base for suitable articles, and known errors are matched with known error and workaround entries. A suggestion has to fit the product and version – if a suggested fix does not match the software version, it is not offered.

That's exactly what the flow is built for. Every value it reads stays linked to the original document, and check rules match it against your ERP. Discrepancies, missing data or unclear matches create a task for the person responsible – with the document, the comparison and decision options. The case then continues right where it left off. Values that aren't in the document are never made up.

You decide: in the XFLOWLY Cloud on a dedicated virtual machine at Google Cloud in Frankfurt – or on your own server (on-premises) running Windows or Linux. Stored data is encrypted with AES-256, and credentials for your systems are additionally kept in a separate secret vault. You'll find all the details in the security factsheet on the platform page.

Easily: after a three-month minimum term, you can cancel XFLOWLY monthly with 14 days' notice. When your subscription ends, we provide a full export – documents, results and your flow definitions in machine-readable formats, available for 30 days at no extra cost. After that, your environment is deleted.

Can’t find your question?Ask us on a call
All questions

More Use Cases

ComplaintsCapture customer and supplier complaints consistently, coordinate actions and track them to closure.Data sheets & certificatesCheck safety data sheets, technical data sheets and certificates, monitor expiry dates and file them properly.Incoming invoicesCapture invoices, match them against the purchase order and goods receipt, approve and post.

Processes that run themselves.

One call, your real process, no prep work. Afterwards, you’ll know what XFLOWLY can do for your business.

Book a call

30 minutes · free · no prep needed

0 cases handled since you got here · Click the logo to send another one through