In May 2024, residents of Supertech Ecovillage 2 in Greater Noida protested late at night after facing repeated power cuts during an extreme heatwave. With temperatures reaching 50°C, families struggled with frequent outages and limited power supply. The incident exposed a growing national challenge: as heatwaves increase and electricity demand rises, India’s power grid is under severe pressure.

⚡ AI Startup Pravāh Aims to Fix India’s Failing Power System

A new AI-driven startup named Pravāh is working to solve this crisis. The company, launched by Stanford alumni Mohak Mangal, Dhruv Suri and Aman Gupta, has gained support from major global investors like Khosla Ventures, Pear VC, and Conviction.
The name Pravāh means “flow”, reflecting the company’s mission to improve the flow of electricity across India.

🔌 Why India’s Power Grid Needs an Upgrade

India’s state-run discoms still depend on outdated forecasting systems, often using spreadsheets and simple data models. These tools fail to predict sudden weather changes, rooftop solar surges, or late monsoons.

When these forecasts go wrong, the consequences include:

• Unexpected power outages

• Costly emergency electricity purchases

• Increased government financial burden

• Higher electricity bills for families

Co-founder Mohak Mangal says, “Discoms cannot rely on old information when energy independence and efficiency are critical.”

🤖 Pravāh’s ‘AI-Native’ Solution for Smart Grid Management

Pravāh has developed a powerful AI-based decision support engine that combines:

• Machine learning demand forecasts

• High-resolution weather modelling

• AI tools for reducing AT&C losses

This technology helps discoms know exactly how much power to buy, when to buy it, and how to manage supply more efficiently. The startup is already working with clients across the US, Germany, and India.

🌧️ AI Weather Research Reveals Global Models Fail for India

The Pravāh team, working with Stanford professor Aditi Sheshadri, studied leading AI weather models and discovered a major problem:
Global AI systems often underestimate extreme rainfall and misread cyclone paths in India.

In a monsoon-heavy country, these errors can be dangerous and costly.

This research now powers Pravāh’s India-specific weather forecasting system, designed to understand local climate behaviour better than any global model.

🔥 Heatwave Maps Show Global Weather AI Frequently Gets India Wrong

The team’s research shows huge mismatches between actual summer heat levels in Gujarat and global model predictions.
If models were accurate, the maps would be white, but they appear blue or red, highlighting errors during severe heatwaves.

This is why Pravāh is developing locally corrected weather models tailored for India’s unique climate.

🔍 AI Gives Grid Operators the One Thing They Always Needed: Foresight

Co-founder Dhruv Suri, who completed his PhD in Energy Science at Stanford, explains:

“AI gives power operators the ability to see the next hour, the next day, and the next extreme event. That’s how you prevent a crisis.”

👨‍💻 A Team Built for India’s Toughest Energy Challenges

The founding trio brings exceptional global expertise:

• Mohak Mangal – Worked with World Bank and MIT Poverty Action Lab across India and Myanmar

• Dhruv Suri – Published work on renewables, built tools at Google X, co-founded clean energy projects

• Aman Gupta – PhD from NYU Courant, collaborated with NASA, IBM, and researched monsoons at LMU Munich and Stanford

🌙 The Bigger Mission: No Family Should Suffer in the Dark Again

The founders say their goal is simple:
To prevent situations like the one in Greater Noida, where parents were forced to carry their children outside at midnight because of long power cuts.

If India can accurately predict storms, heatwaves, and peak electricity demand, families will finally have what every modern household deserves — reliable and uninterrupted power.