WorkOnGrid Raises ₹22.5 Cr To Expand Internationally
WorkOnGrid, a Bengaluru-based AI-native operations platform for utilities, raised Rs 22.5 crore (about $2.4 million) in a funding round led by Transition VC, with participation from Indian Angel Network, announced April 7, 2026. The startup, nicknamed "DiscomGPT" by some customers, unifies smart-meter, sensor, GIS and financial data into one platform that predicts equipment failures and automates utility workflows; in a pilot with Apraava Energy, its AI flagged high-risk distribution transformers before they failed. WorkOnGrid says it already serves 20+ utilities globally and processes data from millions of smart meters. The round, following prior raises of Rs 1 crore in 2020 and Rs 5 crore in 2022, will fund international expansion and deeper AI/ML investment, a concrete example of predictive-maintenance ML reaching production in critical infrastructure.
For ML and data infrastructure practitioners, the interesting part of this raise is not the size of the check, it is where predictive-maintenance AI is actually shipping: a live utility pilot that flagged failing transformers before they broke, not a lab benchmark. The deal shows a recognizable pattern for vertical AI, a narrow, fragmented data problem solved with a domain-specific ingestion and lakehouse layer, then sold as SaaS to operators who cannot build this themselves.
What happened
WorkOnGrid, an AI-native operations intelligence platform for power, water and gas utilities, closed a Rs 22.5 crore (about $2.4 million) round led by Transition VC, an energy-transition-focused fund co-founded by Raiyaan Shingati and Mohammed Shoeb Ali, with participation from Indian Angel Network (IAN), the company said in an April 7, 2026 press release. Bengaluru-based WorkOnGrid, founded by Udit Poddar, Shreyansh Jain, Aayush Agrawal and Shaurya Poddar, had previously raised Rs 1 crore in 2020 from Start Up O Ventures and Rs 5 crore in 2022 from IAN and other angels. The new capital will fund global go-to-market expansion, deepen AI and machine learning capabilities, and build international infrastructure.
Technical context
WorkOnGrid's platform, called Grid, ingests smart-meter, sensor, GIS and financial data into a single data lakehouse, then layers operational intelligence, workflow automation and AI-driven analytics on top, according to the company. That is meant to move utilities from reactive maintenance toward predictive and increasingly autonomous decision-making. In a pilot with Apraava Energy, the platform's AI framework identified high-risk distribution transformers ahead of failure; WorkOnGrid says the approach can generate meaningful annual savings by avoiding outages, though it has not published exact figures. The company says it already serves more than 20 utilities globally and processes data tied to millions of smart meters.
For practitioners
Utility operators sit on fragmented SCADA, GIS, AMI and ERP data that most ML teams never touch, and WorkOnGrid's bet is that a purpose-built ingestion and lakehouse layer is the real unlock, not a bigger model. Teams building predictive-maintenance or anomaly-detection systems in similarly fragmented, low-observability industrial settings (utilities, manufacturing, mining, the adjacent sectors WorkOnGrid also targets) hit the same integration-before-inference problem long before model choice matters.
What to watch
Watch for WorkOnGrid to publish hard numbers on the Apraava Energy transformer-failure results, for utility customers landing outside India as the company deploys this capital toward international expansion, and for whether other grid-analytics vendors respond with similar lakehouse-plus-agentic-AI positioning.
Key Points
- 1WorkOnGrid raised Rs 22.5 crore led by Transition VC to expand its AI-driven utility operations platform internationally.
- 2Its Grid platform unifies smart-meter, GIS, SCADA and financial data into one system that predicts equipment failures early.
- 3A working pilot with Apraava Energy shows predictive-maintenance ML moving from research concept to paid production deployment.
Scoring Rationale
A modest (~$2.4M) vertical-SaaS raise for a single-country utility-AI startup, not industry-shaking, but grounded in a genuine production ML use case (predictive transformer-failure detection) validated across 20+ paying utility customers. Core facts are corroborated by the company's own press release plus three independent outlets (Entrackr, Mercom India, Entrepreneur India), keeping it solidly in the notable-but-narrow band.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
