Inbox classification and routing
AI classifies inbound mail (deadline, status, request, complaint) and routes to the right advisor. Pseudonymized, with full audit log.
Effekt: 2–4 hours per advisor per week saved.
AI for tax advisors
AI for tax advisors — DATEV-aware and §57 StBerG-compliant.
We build AI workflows that handle the inbox, structure correspondence and accelerate research — without violating professional confidentiality.
Three problems we run into
Client data in standard ChatGPT — §57 StBerG violation
§57 StBerGStaff paste financial data and correspondence into ChatGPT for summarization. That breaches the professional duty of confidentiality.
Inbox overload, low-leverage routing
OperationsRoutine inbound mail (deadlines, status updates, document requests) eats hours per advisor per week. AI classification and triage saves the most time, fastest.
DATEV integration uncertainty
DATEVDirect DATEV integrations exist but aren't trivial. We design interface architectures that pre-process AI outputs into DATEV-compatible structures without giving the AI access to DATEV itself.
What we build for tax firms
Document handling and pre-structuring
Receipt images and unstructured correspondence pre-processed by AI, output in DATEV-compatible structure. No AI inside DATEV.
Effekt: Faster month-end close, fewer manual booking corrections.
Tax-law and case-law research
Connection to publisher databases. AI summarizes long judgments, compares clauses, structures search outputs. No client data involved.
Effekt: Initial research in 5–10 min instead of 30–60.
Our approach for tax firms
Workshop
Half day. Use-case prioritization, §57 StBerG risk, DPO involvement.
Pilot
Inbox classification or document handling, pseudonymized, in trial.
Roll-out
Expansion to more workflows, staff training, DPA bundle.
Stimmen unserer LiteLog-Kunden
Direkte Belege, dass Compliance-Software bei uns produktiv läuft – nicht nur in PowerPoint.
„Thanks to LiteLog we could prove we were at the right place at the right time, and that we did our job. Without that proof, we'd have paid 30,000 € in damages."
„The whole tool is well laid out and easy to use. Questions get answered quickly and without fuss. I can fully recommend it."
Frequently asked from tax firms
What can a tax advisor under §57 StBerG actually do with ChatGPT?
Research on tax law and case law is fine as long as no client data is entered. Client correspondence and document handling — only via Azure OpenAI (EU region) with a DPA, or with consistent pseudonymization. Standard ChatGPT with client data violates the professional duty of confidentiality.
Can AI be connected to DATEV?
Direct integrations via official DATEV APIs are possible (receipt images, booking batches). More common is an interface architecture: AI processes receipt images or pre-structures correspondence, then the result flows in structured form into DATEV — pseudonymized, without the AI reaching into DATEV itself.
Which use case pays off fastest?
In practice: classification of incoming mail and routine client correspondence. Both are manageable from a data-protection standpoint and save several hours per employee per week. Tax-related research is a good second use case.
What about the Chamber of Tax Advisors — are there already rules?
There are currently no chamber-mandated technical rules on AI use. The professional principles (confidentiality, due care, personal responsibility) still apply unchanged — we help you document your architecture so it withstands future rules too.
How is the firm's data protection officer (DPO) involved?
We draft the DPIA together with the DPO, provide DPA templates and technical/organizational measures. If you don't have an external DPO, we can recommend one.