United StatesTechnology and telecom2025 – ongoingIBM
AI tools for log analysis and routine case tasks in software support
IBM's software support engineers saved an average of 40 minutes a day after AI automated routine case tasks
Problem
IBM's software support engineers handled large volumes of cases. Many were routine, and many required reading log files by hand.
Approach
IBM built three tools on watsonx.ai and watsonx Orchestrate. One classified log files and flagged anomalies, one automated follow-ups, updates and closures, and one answered questions from documentation. The automation tool launched across 14 use cases with 100 engineers and was then extended to 2,800 engineers.
Results
Takeaway
The measured gains came from log analysis and routine task automation. The documentation answer tool had a stated accuracy of 28%, so its time saving held only when answers were right.
Categories
- Size
- Large enterprise
- Industry
- Technology and telecom
- Business function
- Customer serviceIT and R&D
- Innovation type
- Process innovation
- AI technology
- Generative AIAI agentsPredictive analytics and ML
- Adoption
- Built in-house
- Evidence
- Vendor case study
Sources
Source summaries
IBM describes its own use of watsonx tools in software support. It reports 40 minutes saved per engineer per day from automated case tasks and 31 minutes saved per case from log analysis. It also gives a 28% accuracy figure for its documentation answer tool.
Full details are in the original sources.