Paper forms and clipboards
Reports, checklists and orders are filled in on paper on site and typed up later in the office. Two entries, one result, errors in between.
Process automation and AI consulting from Berlin: we take the manual work out of your workflows, from digital capture to autonomous case handling. AI agents take over incoming mail, checking and posting and hand tasks on to each other, while your staff decide at the approval points. Plus digital forms, document management, the Power Platform, interfaces and speech capture. The process comes first, then the tool.
Most slow processes share the same causes. If you recognise one of these points, that is the process we start with.
Reports, checklists and orders are filled in on paper on site and typed up later in the office. Two entries, one result, errors in between.
A case switches several times between paper, PDF, email and the business application. Every switch costs time and traceability.
Data from emails, invoices or orders is typed by hand into ERP, CRM or accounting, often more than once.
Leave, invoice, quote: the approval sits in a mailbox, nobody sees the status, follow-up questions travel down the corridor.
Important stocks, dates or customer data live in spreadsheets circulated by email, in several versions and without permissions.
Letters, faxes and email attachments are reviewed, assigned and filed by hand. What matters is decided by whoever has time.
AI agents take over clearly defined tasks end to end: picking up requests from email, post and portal, creating cases, fetching data from your business systems, checking it against rules, preparing replies and handing the task on to the next agent. Every agent has clearly limited rights, every action is logged, and nothing leaves the company without approval by your staff.
Incoming invoices, delivery notes and remittance advices are recognised, matched against orders, pre-coded and handed over to financial accounting via an interface. The approval path stays, but runs digitally and by rules, for example by amount or project. Your accounting team reviews exceptions instead of every invoice.
Microsoft Copilot in Microsoft 365, analysing and summarising documents, classifying incoming mail, preparing replies, reading data from receipts. Beforehand we clarify data access, permissions and data protection; without that clarification we do not start.
Paper and PDF forms become digital forms with mandatory fields, photos, signature and plausibility checks, offline as well. One example is our own application eMODAT; depending on the environment we also use Microsoft Forms, Power Apps or forms inside your business application.
Scanning, text recognition (OCR), automatic classification and filing in SharePoint or your document management system. Invoices, contracts and mail arrive with metadata where they belong and can be found.
Approvals, notifications, tasks and status tracking as workflows in the Microsoft world, connected to Teams, Outlook and SharePoint lists. Small apps for procedures that no standard software covers.
Data flows without retyping between ERP, CRM (for example vTiger), DATEV, Microsoft 365 and your business applications. Via standard interfaces where they exist, otherwise via our own connectors from our software development.
Where text is created, speaking is faster than typing: findings, reports, minutes and notes by speech recognition straight into the application, at the desk or on the move. One example is our product ProDictate.
A typical entry point. We start with a process that noticeably ties up time today, collect the quick improvements first and scale from there.
Document the workflow, the people involved, documents, systems and waiting times, the way the process really runs today.
Find the points that can be improved right away with existing means, such as a form, an approval or an interface.
Pick the right means from the toolkit, sketch form, workflow, interface or AI agent, define the approval points and walk it through with the users.
One department or one site uses it in production, feedback flows straight back in.
Further processes and sites follow the same pattern, with the lessons from the pilot and key figures from live operations.
Anonymised project examples from our work: an AI-supported service charge statement, an accounting process that runs from invoice receipt to posting without a media break, and AI agents that prepare cases in customer service. All examples with starting point, approach and result are on the project examples page.
Not always. A large effect often comes from digital forms, clear workflows and interfaces. AI is added where unstructured content has to be processed or decisions have to be prepared, such as mail, receipts, unstructured text or entire cases in case handling.
An AI agent takes over a clearly defined task from start to finish: it picks up a case, fetches the necessary information from your systems, checks it against defined rules, prepares the result and hands it on to the next step, to another agent or to a person. Which decisions it makes itself and where an approval is required is agreed with you in advance. Every action is logged.
Large parts of it: invoice receipt, recognition, matching against orders, pre-coding, digital approval and handover to financial accounting run without manual entry. Approvals and exceptions stay with your staff, and responsibility for the year-end close stays with accounting and your tax advisor. How far we go depends on your documents, systems and audit obligations.
No. eMODAT and ProDictate are two tools among several. Often what you already license with Microsoft 365 is enough, such as Power Automate, Power Apps and SharePoint. We choose the means according to the process, not the other way round.
Before every deployment we clarify which data is processed, where it resides and who may access it. Without that clarification we do not start.
Usually yes. ERP, CRM such as vTiger, DATEV and Microsoft 365 come with interfaces. Where one is missing, we develop a connector. Which systems are affected in your case is clarified during the process capture.
That depends on complexity and interfaces. We deliberately keep the first process manageable in scope and tell you the effort after the capture phase.
We start with the workflow that currently ties up the most time.