ItalyFinance and insurance2025 – ongoingUnipol
NAMI AI platform for monitoring and resolving IT system events
Unipol's AI platform NAMI cut incident handling time by 90%
Problem
Unipol ran more than a dozen separate monitoring systems. Only 26% of system events were covered, and staff handled incidents by hand.
Approach
Unipol built NAMI, an automation platform on IBM watsonx, with generative AI use cases and agentic workflows run through watsonx Orchestrate. It uses several models, including Llama, Mistral's Pixtral and IBM Granite. The monitoring systems feed one data lake, and NAMI analyzes events and resolves many of them on its own.
Results
Takeaway
Unipol put the AI on event monitoring, where each incident was small and the volume was high. Staff were left with the cases NAMI escalated.
Categories
- Size
- Large enterprise
- Industry
- Finance and insurance
- Business function
- IT and R&D
- Innovation type
- Process innovation
- AI technology
- AI agentsGenerative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
IBM describes how Unipol built the NAMI platform on watsonx with agentic workflows and several language models. The page reports monitoring coverage rising from 26% to 100% and incident handling time dropping by 90%. In two months NAMI analyzed over 800 events and escalated 538.
Full details are in the original sources.
