A power trading desk in India now makes hundreds of decisions a day: buy or sell in the Day-Ahead Market, top up in the Real-Time Market an hour before delivery, manage green obligations, hedge, and keep the schedule close enough to actuals to avoid deviation charges. Each decision depends on data that has to be right before a gate closes.
What the market looks like
India's power market in FY2025-26
Market data, not DaasLabs results
| Measure | Volume |
|---|---|
| Total electricity transacted | 1,707 BU |
| Short-term market | 302.08 BU (17.7%) |
| Exchanges: IEX / PXIL / HPX | 141.14 / 18.35 / 9.06 BU |
| Bilateral | 96.04 BU |
| Deviation settlement | 37.49 BU |
Note: Source: Indian Infrastructure, July 2026.
Source: Indian Infrastructure, “Trading trends: shift to market-based solutions” (2026)
The year before, short-term transactions reached 238.35 BU, up 9.22%, and REC trading rose to 30.42 million certificates from 11.6 million (Power Line). The direction is clear: more power, and more of it priced in markets.
The rules are moving
- July 2025: CERC ordered phased market coupling, starting with the DAM; monthly baseload derivatives launched on MCX and NSE after SEBI approval.
- December 2025: guidelines for virtual power purchase agreements.
- April 2026: a draft amendment proposes Grid Controller of India as Market Coupling Operator, for one uniform clearing price.
- May 2026: deviation settlement moves to daily weighted-average pricing.
Where data and agents help the desk
A supervised trading-desk squad
Agent designs from our AI & Agentic Engineering practice
| Agent | What it does | Who decides |
|---|---|---|
| Price & Load Forecaster | DAM / RTM price and load forecasts by block | Head of trading analytics |
| Bid Preparation Agent | Drafts bids and balances the portfolio | A trader submits every bid |
| DSM & Imbalance Watch | Flags blocks heading for deviation charges | Scheduling desk |
| Settlement Reconciler | Matches settlements, explains breaks | Accounts officer, beyond tolerance |
Note: Designs, not results.
Where to start
Start with one portfolio: publish load, price and RE forecasts as data products with a freshness SLA tied to DAM bidding, add DSM watch and settlement reconciliation, and measure power-purchase cost and deviation charges against today's baseline.