Hey, Lumi! Using Natural Language for Intent-Based Network Management
summary
Builds a working system called Lumi that turns freeform operator requests into a structured intent language (Nile), compiles them into device configs (targets merlin and extends it for quotas and temporal rules), and deploys changes in milliseconds. Evaluations use a 150-item synthetic set and 50 real campus intents, report high precision/recall for entity extraction, show feedback reduces errors over rounds, a 26-person user study with most tasks completed, and an 88.5% user preference for the interface.
AI-written summary · about
chapters (5)
- 0:00Introduction to Lumi
- 1:37System design and features
- 3:34Implementation details
- 7:12Evaluation and data analysis
- 10:52User study and conclusions
related talks
11mAI-Native Network Operation
40mNew Network Provisioning System Leveraging Kubernetes and Cloud Native Open Source
38mNmstate Polyglot Model - Translating Natural Language into Nmstate States
18mMaking Logic a First-Class Citizen in Generative ML for Networking
12mAI-Native Network Operation
16mUNUM: A New Framework for Network Control
29mSwitchLog: A Logic Programming Language for Network Switches
24mNet2Text: Query-Guided Summarization of Network Forwarding Behaviors