The short answer: In 2026, law firms typically automate three areas around RA-MICRO and DATEV: client intake (form → file creation), the beA inbox (capturing and classifying incoming messages, deriving deadlines) and deadline control (calendar, advance deadlines, escalation). The core legal decision – such as recording the deadline itself – stays with the licensed lawyer; what gets automated is data handover, notifications and routine tasks via middleware such as Make or n8n, connected to the APIs of RA-MICRO, DATEV and beA.
The key point: it's not beA sending or the binding deadline calculation that gets "automated away", but the manual transfer between systems, the copy-pasting, the searching and the remembering.
What can sensibly be automated – and what can't
Well suited to automation:
- Client intake from the web form through to structured file creation
- Capturing and classifying beA messages (sender, case reference, message type)
- Advance-deadline and escalation notifications to the responsible department
- Document filing and naming conventions in the DMS/file structure
- Status sync between the practice management software and a client portal
- DATEV handovers for billing and accounting (RVG/time → invoice)
Deliberately NOT fully automated:
- The binding recording and calculation of deadlines – the four-eyes/licensed-lawyer check remains mandatory here.
- beA sending of deadline-preserving briefs – approval and qualified signature/trusted origin remain manual.
- Substantive assessment of incoming mail.
Automation takes over the groundwork, not the lawyer's responsibility. Exactly this dividing line is what makes a system fit for a law firm.
The stack: RA-MICRO, DATEV, beA and the automation layer
In practice, an automated law firm workflow consists of four layers:
- 1Core systems (system of record): RA-MICRO (files, deadlines, conflict checks) and DATEV (accounting, payroll, billing). These remain the leading systems – nothing is replaced here.
- 2beA connection: access to the special electronic lawyer mailbox (beA) via the interface provided by the BRAK (German Federal Bar) or compatible practice software modules. Messages are read and passed on as events.
- 3Automation layer: Make (visual scenarios, fast implementation) or n8n (self-hosted, more control, EU hosting/on-premise). This layer orchestrates: it listens for events, transforms data and calls target systems.
- 4Interface/portal: a Softr portal or dashboard for clients and the team – for intake, status queries and document uploads, without clients getting access to the practice software.
The APIs at a glance: RA-MICRO provides access to file data via RA-MICRO Online / interfaces; DATEV offers connections for receipts and master data with DATEVconnect and the DATEV API; beA access runs through the official BRAK interface. Where a direct API is missing, we work with structured exports, watch folders or RPA as an interim solution.
Workflow 1: Client intake without duplicate data entry
The most common time sink is typing client data in multiple times. A clean intake flow looks like this:
- 1Capture: the client fills out a Softr/web form (name, matter, opposing party, document upload). Required fields enforce completeness.
- 2Conflict pre-check: Make/n8n checks the names of client and opposing party against the file base and flags potential conflicts of interest for manual approval.
- 3File creation: after approval, the data is passed to RA-MICRO in structured form (file creation, parties, case handler).
- 4Documents: uploaded files land in the file structure with a consistent naming convention.
- 5Confirmation: the client automatically receives an acknowledgement and – once the mandate is accepted – access to the portal.
Effect in practice: instead of 15–20 minutes of manual setup per new matter, all that remains is an approval click of under a minute. With 40 new matters a month, that's around 10 hours gained – time previously lost in the front office.
Workflow 2: Processing the beA inbox in a structured way
The beA inbox is the most critical entry point for deadlines. Manual review is error-prone and slow. An automated intake process:
- 1Capture: new beA messages are detected via the practice software connection.
- 2Classification: sender (court, opposing party, authority), case reference and message type are extracted – for unstructured attachments, AI text extraction can provide suggestions (e.g. "order with deadline detected").
- 3Assignment: the message is assigned to the correct file in RA-MICRO via the case reference.
- 4Deadline suggestion: if the system detects a potential deadline, it generates a suggestion with an advance deadline – for review, not for automatic recording.
- 5Notification: the responsible department receives a structured alert (Teams/email) with a link to the file.
Important: the AI provides a suggestion; a human confirms the deadline entry. This combines speed with professional diligence. Incoming mail becomes transparent within minutes instead of being reviewed in a batch at the end of the day.
Workflow 3: Deadline management with advance deadlines and escalation
RA-MICRO keeps the deadline calendar – that stays the same. What gets automated is the communication around it:
- Advance-deadline reminder: an automatic alert to the case handler and lawyer X days before expiry.
- Escalation levels: if an advance deadline isn't marked as "handled", the notification escalates to the department head.
- Daily digest: every morning, a consolidated overview of open deadlines per team – one view instead of many calendar clicks.
- Weekly report: key figures for firm management (open deadlines, completion rate, anomalies).
Automation here is a safety net, not a replacement for deadline control. It ensures no one misses an advance deadline because a message got lost in an inbox.
No-code vs. custom: the decision framework
The central build question is: where is no-code enough, and where is custom development needed?
No-code/low-code (Softr + Make/n8n) fits when:
- Standard processes with clear rules are being automated (intake, notifications, status sync).
- You want to go live quickly and iterate (weeks instead of months).
- The team should be able to adapt the solution themselves later.
- The connection is feasible via existing APIs/exports.
Custom development pays off when:
- Complex, firm-specific logic or deep integrations are required.
- High scale, special security requirements or an on-premise mandate apply.
- A standard platform hits clear limits.
Our recommendation for most law firms: start with a no-code core that covers 80% of the routine, and develop only the remaining 20% as targeted custom work. n8n is often the better choice over pure SaaS because it can be run EU- or self-hosted – a relevant argument for law firm data. One principle is decisive: the firm must own the system – documented, transferable, with no dependence on a single person.
Implementation in four phases
- 1Process audit (weeks 1–2): mapping the current state of the three core flows, measuring manual effort, clarifying interfaces (RA-MICRO, DATEV, beA).
- 2Pilot (weeks 3–6): one workflow – usually intake – is implemented end to end and tested with one department.
- 3Rollout (weeks 7–12): beA intake and deadline notifications follow, including training and documentation.
- 4Operations & ownership: handover to the team with monitoring, error alerts and a short operating guide. The firm can adjust rules itself.
Realistic timeframe: a clean first rollout of the three workflows is productive in 8–12 weeks – provided API access has been clarified.
Compliance and data protection, not as an afterthought
Law firm data is subject to attorney confidentiality (§ 43a BRAO) and the GDPR. For automated workflows, this means specifically:
- Prefer EU hosting/on-premise (n8n self-hosted, EU data centers).
- Data processing agreements with all services used.
- Data minimization: send only the fields needed for each step through the automation.
- Strictly separate access roles – client portal ≠ file access.
- Logging: log every automated action traceably.
- AI use only with contract-compliant, EU-hostable models and without unnecessary transfer of matter content.
The basic rule remains: automation reduces human error, but does not replace ultimate professional responsibility.
FAQ
Can beA sending be automated?
Deadline-preserving sending should not be fully automated. What can sensibly be automated is the inbox (capturing, classifying, assigning) and the preparation of outgoing items; final approval and sending remain with the licensed lawyer.
Are there official interfaces to RA-MICRO and DATEV?
Yes. RA-MICRO provides access to file data via RA-MICRO Online and modules, DATEV via DATEVconnect and the DATEV API. Where a direct API is missing, we work with structured exports, monitored folders or RPA. Availability depends on license and version and is checked during the audit.
Make or n8n for a law firm?
For maximum data control and EU/self-hosted operation, n8n is usually the better choice. Make scores with fast implementation and many ready-made connectors. Often the two are combined – or firms start with n8n as a privacy-friendly core.
How long does implementation take?
A pilot workflow is often productive in 3–6 weeks, the full rollout of the three core flows in 8–12 weeks – depending on clarifying API access and internal approvals.
Does the firm keep control of the system?
Yes – that is the central requirement. Workflows are documented, equipped with monitoring and handed over to the team, so rules can be adjusted without depending on a single person.
Want to cleanly automate intake, beA intake and deadline notifications – GDPR-compliant and fully owned by you? Mindflows builds these systems on Softr, Make and n8n for law firms in the DACH region. Get in touch for a process audit.