Grid software: the layer India's energy transition has not yet built

For most of its history the grid has been a hardware system: generation, transmission and the maintenance of frequency. Three changes in India over the last three years have made it a software problem as well, and one that a seed-stage company can address.

Three shifts

Storage became economic. In June 2025 the Ministry of Power approved viability gap funding of ₹5,400 crore for 30 gigawatt-hours of standalone battery storage, at ₹18 lakh per megawatt-hour, with a minimum of 20 percent domestic content. That is projected to mobilise about ₹33,000 crore of private investment, and it creates, for the first time, a policy-protected market for Indian storage components. Solar with storage is now being discovered at tariffs that compete with coal for the evening peak.

Distribution became a data business. Smart meters installed under the distribution reform scheme reached 5.28 crore in early 2026 against 20.33 crore sanctioned, and the roll-out is the infrastructure build. Every meter is a sensor. Aggregate technical and commercial losses fell from 21.91 percent in FY21 to 15.04 percent in FY25, which means loss reduction is a monetisable service rather than an aspiration.

Source: Ministry of Power, Revamped Distribution Sector Scheme progress data, 2026.

Source: Ministry of Power, Revamped Distribution Sector Scheme review.

Vehicles became grid assets. A rapid-charging network draws power the way a small factory does; a fleet of swappable packs is a distributed battery bank. Exponent's chargers, Pulse Energy's network and Sheru's energy bank sit at the boundary between mobility and the grid, and the data they produce is the raw material for everything below.

The missing layer

Above the hardware sits orchestration: forecasting load and renewable output, scheduling, dispatching storage, aggregating distributed resources into a virtual power plant, trading power in the day-ahead and real-time markets. In other markets this is where the software companies sit. In India it is almost empty at seed stage. The reasons are practical. Access to smart-meter data is not yet standardised. State distribution utilities are difficult counterparties, slow to buy and slower to pay. And grid software budgets in India are small today, so a company has to find a buyer who is not a utility.

That buyer exists. Commercial and industrial consumers on open access, renewable developers who must schedule and forecast under deviation penalties, traders, storage operators, charging networks and the new class of data-centre loads all pay for software that reduces their cost of power. The distribution utility can come later, when the product already works.

What we would back and what we would not

We would back a forecasting and scheduling product with paying customers among renewable generators and traders; a virtual-power-plant operator built on assets someone else owns; battery analytics that predict thermal events, which we expect to become a condition of storage funding by FY28; and second-life storage built on a chemistry-agnostic diagnostic. We would not back storage integration, where Tata, Reliance and the Chinese integrators leave no right to win for a seed cheque, or project development, or a software company whose revenue plan depends on a single state utility.

We came to energy through mobility, and the portfolio gives us demand-side data on how vehicles draw power. The next few energy investments will be software, diagnostics and components. The grid is being rebuilt, and the software that runs it has not yet been written.

Sources: Ministry of Power and MNRE releases on viability gap funding for battery storage (June 2025); Revamped Distribution Sector Scheme progress data (2026); Ministry of Power review of AT&C losses; AdvantEdge portfolio data.