Case Studies / Tirol Kliniken Innsbruck
HEALTHCARE · HOSPITAL OPERATIONS
Automating patient data anonymization with AI at Tirol Kliniken
Tirol Kliniken – the largest healthcare provider in western Austria – needed to anonymize thousands of medical documents to meet regulatory requirements. Staff were doing it by hand. We deployed ShareMedix, our AI-powered anonymization platform, to automate the entire process on their local servers.
Client: Tirol Kliniken GmbH – the largest healthcare company in western Austria, providing comprehensive medical care across multiple facilities in Tyrol.
KEY RESULTS
100%
Manual anonymization effort eliminated
TGF
Full compliance with Tyrolean Health Fund standard
On-prem
Deployed on hospital’s local servers – no cloud dependency
No GPU
Runs on standard hospital infrastructure without GPU
INDUSTRY
Healthcare
USE CASE
Medical document anonymization
AI APPROACH
NLP + continuously learning AI
DOCUMENTS TYPES
Discharge reports, doctors’ letters, findings
The challenge
Tirol Kliniken generates a large volume of medical documents every day – hospital discharge reports, doctors’ letters, findings reports, and other clinical records. Many of these documents need to be anonymized before they can be shared, archived, or used for secondary purposes, in compliance with the TGF (Tyrolean Health Fund) standard.
Until this project, anonymization was done manually. Staff reviewed each document, identified patient-identifying information, and redacted it by hand. The process was slow, expensive, inconsistent, and didn’t scale with growing document volumes and increasing regulatory demand.
The core problem: a growing volume of medical documents required anonymization to meet regulatory standards, but the manual process was too time-consuming and costly to keep up – and the inconsistency created compliance risk.
What we built
We deployed ShareMedix – our AI-powered medical document anonymization platform – at Tirol Kliniken. The system automates the entire anonymization workflow, from identifying patient data in clinical documents to rendering it unrecognizable.
Intelligent data recognition. The AI algorithms were trained to identify and classify patient-identifying information across different medical document types – names, dates, addresses, insurance numbers, and other personal data – while preserving the clinical content that needs to remain visible and usable.
Medical terminology handling. Healthcare documents are dense with specialized terminology. The models were specifically trained to distinguish between clinical terms and personal identifiers, avoiding false positives that would strip out medically relevant information.
On-premise deployment without GPU. Given the sensitivity of patient data, the system had to run entirely on the hospital’s local servers — no data leaving the premises. A key technical constraint: the solution needed to work on standard hospital hardware without GPU acceleration, which required careful model optimization.
Continuous learning via user interface. Hospital staff interact with ShareMedix through a web-based interface where they can review anonymized documents, make corrections, and add notes. These corrections feed back into the model, allowing it to improve over time on the specific document patterns and formats used at Tirol Kliniken.
The results
BEFORE
Manual anonymization of every medical document. Slow, expensive, inconsistent quality, and difficult to scale with growing regulatory requirements.
AFTER
Automated anonymization in compliance with TGF standards. Staff review and correct edge cases through an intuitive UI, and the system learns from each correction.
The manual anonymization workload was eliminated. Documents that previously required individual staff attention were now processed automatically, with human review only needed for edge cases. Compliance with the TGF standard was maintained consistently across all document types.
The continuous learning mechanism means the system gets better with use, each correction made by staff improves future accuracy, creating a feedback loop that reduces the need for human intervention over time.
Technology used
More Case Studies
See how we’ve helped other companies
AUTOMOTIVE (UNDER NDA)
Replacing reactive maintenance with a predictive maintenance roadmap
Unplanned equipment failures were causing costly production stops. We assessed the manufacturer’s infrastructure and delivered a concrete architecture for predictive maintenance, a roadmap from reactive repairs to data-driven prevention.
PdM
Architecture
Roadmap
Delivered
MANUFACTURING · RADAWAY
Making email-based order processing reliable with LLMs
Staff were manually reading customer emails, identifying products, and entering orders by hand. We turned a promising AI prototype into a production system that handles it end to end, across languages, formats, and attachments.
-90%
Manual intervention
95%+
Match accuracy
LOGISTICS · FR. MEYER’S SOHN
Eliminating manual data extraction from thousands of daily shipping emails
Operations staff were manually reading German and English logistics emails to pull out routing and scheduling data, every single day. We built an AI pipeline that extracts, structures, and delivers the data automatically.
–80%
Manual effort
2 langs
DE & EN
On-prem
Deployed