IrelandHealthcare and life sciences2025 – ongoingMedtronic
Document Intelligence Platform for invoice and delivery matching
Medtronic cut invoice matching from up to 20 minutes to 8 seconds with a document AI platform
Problem
Medtronic staff matched invoices, proofs of delivery and purchase orders by hand. Stamped, poorly scanned and non-standard documents made the work slow.
Approach
Medtronic worked with IBM Consulting and IBM Neudesic on a Microsoft-based Document Intelligence Platform Accelerator. It uses machine learning and generative AI to read and match the documents. Automated matching went live in one department within four weeks.
Results
Takeaway
Document matching is a good fit for AI when the files are messy. Staff time moves to higher-value work once routine matching is automated.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Finance, risk and legal
- Innovation type
- Process innovation
- AI technology
- Generative AIPredictive analytics and ML
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
IBM describes how Medtronic automated matching of invoices, proofs of delivery and purchase orders with machine learning and generative AI. It reports processing time falling from up to 20 minutes to 8 seconds per invoice and accuracy above 99%. It says the first department went live in four weeks.
Full details are in the original sources.
