Grid intelligence that runs where the grid does.

GridVeda monitors power transformers in real time using three AI models on NVIDIA hardware. No cloud. No delay. Just answers.

What you get

Three views. One system.

Fleet Overview

Monitor all 20 transformers at a glance with real-time health scores and AI alerts.

GridVeda fleet overview dashboard showing 20 transformer cards with health metrics and AI alert feed

Transformer Detail

Deep-dive into individual units -- temperature, DGA, load, and fault history.

GridVeda transformer detail view showing XFMR-020 with temperature, load, DGA charts and live telemetry

Live Grid Load

Real-time grid load ticks and spotlight analysis on high-risk transformers.

GridVeda live grid load bar chart and spotlight panel with temperature and dissolved gas analysis

Under the hood

Three models. One verdict.

Each model handles a different dimension of grid intelligence -- from diagnostics to global awareness to responsible onboarding.

Video DemoYouTube

Conversational Analysis

Nemotron

Analyzes operator screens, translates natural language queries into diagnostics, and explains faults using IEEE standards.

Real-World Intelligence

Perplexity Sonar

Analyzes global grid situations, retrieves past anomaly data, and provides diagrams showing where grid failures will happen based on historical patterns.

Responsible AI Training

oss

Onboards new engineers, teaches them about the system, ensures knowledge transfer and responsible AI practices.

Performance

Built for the edge.

8.3s
Avg detection time
98%
Normal classification
<200ms
End-to-end latency
$249
Edge deployment cost
ComputeNVIDIA RTX 5090 (24GB GDDR7, 10,496 CUDA cores)
Edge TargetJetson Orin Nano Super (67 TOPS, 25W)
BackendFastAPI + Uvicorn with WebSocket streaming
FrontendHTML5 + JavaScript with Web Speech API
ProtocolsModbus, DNP3, IEC 61850, OPC-UA
StandardsIEEE C57.104, Duval Triangle, Rogers Ratio

Research

Read the paper

GridVeda: An AI-Driven Grid Intelligence System for Real-Time Transmission Reliability Monitoring

GridVeda_Research_Paper.pdf
Loading paper...