My Work
Spectrum Defragmentation
Transforming backend-dependent spectrum planning into a real-time visual editing system.
Enterprise network engineers were forced into backend scripting for spectrum adjustments, causing structural workflow friction and repeated failures.
I led end-to-end redesign from discovery workshops to stakeholder alignment and rollout within 3-week window.
Impact
-
70% retry reduction
-
40% faster execution
-
Removed backend dependency
-
Adopted across 3 internal engineering teams
Alarm Storm
ML-assisted monitoring system that helps network engineers detect and diagnose alarm storms faster.
Transforming overwhelming alarm storms into a structured, ML-assisted investigation workflow.
​
During major network outages, operators receive thousands of alarms within minutes, making root cause identification extremely difficult.
Engineers had to manually scan alarm streams and correlate events across multiple nodes, slowing incident diagnosis.
​
I designed an ML-assisted workflow that detects alarm storm segments, groups alarms into clusters, and surfaces potential root causes.
Impact
-
Shifted analysis from alarm scanning to pattern-based investigation
-
Reduced cognitive load during alarm storms
-
Enabled easier investigation of root causes in a single incident
Upcoming Works
ViNCA - AI Assistant Tool
Introducing a conversational layer to enterprise network systems
Designing an AI-powered assistant that enables network engineers, maintenance teams, and leads to interact with complex systems through natural language.
​
From clearing alarms with a single prompt to retrieving network service data instantly, the assistant reduces operational friction and enables workflow automation through custom agents.
​
Seamlessly integrated across the Virtuora product suite.
Case study coming soon.