United StatesTechnology and telecom2025 – ongoingLumen Technologies
Generative AI recommendations and ticket summaries in the Laser diagnostic tool
Lumen Technologies saved an estimated 3,675 hours a year by adding generative AI to its technician diagnostic tool
Problem
Lumen's engineers worked with alarms, performance metrics and ticket notes stored in separate systems and formats. Service tickets could run 15 to 20 pages. Its Laser diagnostic tool, which handles over 3 million diagnostics a year, relied on logic about five years old.
Approach
Lumen added generative AI to Laser on Azure Databricks. Structured inputs go through model serving to a Llama 3.3 model hosted on Databricks, and technicians start the AI step with a button in the tool. The AI also summarizes the last 90 days of a customer's ticket history so technicians can get up to speed quickly. Lumen wrote its prompts in the Agent Bricks AI Playground and later added Unity Catalog for governance.
Results
Takeaway
Putting AI answers inside a tool technicians already used meant no new workflow to learn. Lumen's time saving is an estimate built on a per-task average.
Categories
- Size
- Large enterprise
- Industry
- Technology and telecom
- Business function
- Customer serviceProduction and quality
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks reports that Lumen added generative AI recommendations and ticket summaries to its Laser diagnostic tool. It gives an estimate of 3,675 hours saved a year, based on about 15 minutes saved per task. Lumen aims to raise ticket deflection from 30% to over 50%. The page gives no launch date.
Full details are in the original sources.