FaultLine
Network-fault diagnosis: a rule engine and an AI, judged by a human
- Status
- Internship project
- Year
- 2026
- Rooms
- AI & Agents, ML & Impact
Problem
Manual network troubleshooting is slow and inconsistent.
Approach
A deterministic Python rule engine and an AI model read identical evidence from a Packet Tracer topology; findings are shown side by side and a human accepts, edits or rejects each AI recommendation before the fix is applied and re-tested.
Engineering decisions
- Human-in-the-loop review is logged as evidence of governance.
What it does
- Six fault types
- Parallel rule vs AI diagnosis
- Accept / edit / reject review
- Verification by re-test
Stack
My part
Built during the Cisco × AICTE virtual internship in AI; 13 of 15 commits are mine. This was an internship programme, not employment at Cisco.
Limits, stated plainly
- Evaluated on injected faults in a simulated topology.
Evidence
- CISCO-Project README
- Cisco AICTE internship certificate
Screenshots
FaultLine running the full pipeline: rule engine and AI agreement on injected faults
FaultLine dashboard comparing rule-engine and AI diagnoses across cases