GE Digital Ghost - Embedded IoT & Kibana Analytics

Stabilized GE's Digital Ghost C++ system and built custom Kibana React plugins for real-time Elasticsearch data access and advanced analytics

Client ·
ChallengeImprove embedded system stability and big data observability for Digital Ghost
ResultResolved memory issues and created Kibana plugins to boost export and analytics by 30%
TagsBig Data, JavaScript & TS, C++ & Embedded, Public, Energy & Transport

General Electric Global Research partnered with TaylorMade Software to enhance the performance and observability of their Digital Ghost embedded monitoring system, used for real-time power grid telemetry in high-demand environments like NYC's Con Edison.

We identified and resolved persistent C++/C memory leaks in GE's proprietary memory pool libraries using Valgrind and GDB, stabilizing long-running embedded workloads.

On the analytics front, we developed a custom Kibana plugin in React and Node.js that streamed in-memory data and exported Elasticsearch datasets using elasticdump, via an intuitive Query-By-Example (QBE) UI.

Our enhancements to Kibana's UI and service layer enabled new index field creation with metadata tagging, which significantly improved search, filtering, and dashboard insights by 25%.