In a power grid, a forecast is a financial instrument. If your solar farm over-promises and under-delivers, you pay. For a new Delhi startup, the bet is that a better forecast, built from space, can turn those penalties into profit.
BioSky Space Innovations, an IIT Delhi spinout founded in early 2025, is selling satellite-derived intelligence to renewable energy producers. The company claims its platform, which fuses multi-sensor Earth observation data with proprietary AI models, can cut the demand-side-management penalties for its live clients by roughly 50% [BioSky, Oct 2026]. For an industry where margins are measured in single-digit percentages, that is a number that gets attention.
The physics-informed wedge
Most renewable forecasting leans heavily on global weather models. BioSky’s pitch is that for the complex, localized climates of the Global South, you need a different lens. The company says its models are physics-informed, trained not just on historical patterns but on the underlying atmospheric and terrain physics specific to a region [Perplexity Sonar Pro Brief]. The output is a forecast designed to inform very concrete, very expensive decisions: when to charge or discharge a grid-scale battery, and how to bid generation into the day-ahead and real-time electricity markets [BioSky, Oct 2026].
The company’s early traction suggests someone is buying the argument. Its website names four live clients: ReNew Power, Serentica, Sembcorp, and Resolven [BioSky, Oct 2026]. These are not pilot projects with municipal utilities; they are some of the largest independent power producers and energy traders in India.
ReNew Power | 1 | Client
Serentica | 1 | Client
Sembcorp | 1 | Client
Resolven | 1 | Client
A team built for the intersection
The founding team appears engineered for the niche where space data meets energy markets. Co-founder and CEO Saurabh Kapil holds an MBA from IIM Calcutta and spent nearly a decade in consulting, including time in PwC’s space practice [APCTT, Mar 2026][The Rocket Media, 2026]. The technical co-founder, Nimish Nama, is not as publicly detailed, but the broader team roster fills out with specific expertise.
| Role | Name | Key Background |
|---|---|---|
| Co-founder & CEO | Saurabh Kapil | IIM Calcutta MBA, PwC Space Practice |
| Co-founder | Nimish Nama | Co-founder listed in company materials |
| Remote Sensing Scientist | Abhishek Yadav | PhD from IIRS/ISRO |
| ML Scientist | Archana Yadav | BS+MS in Data Science, IISER Bhopal |
| ML Engineer | Hemraj Chakravarti | PGDBA from IIM Calcutta |
This blend of commercial space experience, deep remote-sensing pedigree from the Indian Space Research Organisation (ISRO), and machine-learning talent is the kind of resume a venture investor looks for in a deep-tech climate bet. The company is still small, reported as having between one and ten employees [Crunchbase, Retrieved 2026], and is actively hiring for founding engineering roles in AI and power systems [BioSky, Oct 2026][JobLeads, 2026].
Where the forecast could turn cloudy
The ambition is clear, but the path is lined with hard, technical sales. BioSky is entering a field where incumbents range from giant weather-data providers to sophisticated in-house analytics teams at the very utilities it wants to sell to. Its claimed 50-70% reduction in potential revenue loss for producers is a company-associated claim, not yet backed by public, third-party case studies [Perplexity Sonar Pro Brief]. The sales motion here is high-touch, high-stakes, and slow. Convincing a risk-averse plant manager to trust a new startup’s algorithm over their existing toolkit is a formidable task.
The company’s financial footing is also in the earliest stages. Corporate registry data shows an authorized share capital of ₹25 lakh and a paid-up capital of just over ₹2.15 lakh [Perplexity Sonar Pro Brief]. It has garnered non-dilutive support, including up to ₹50 lakh from the AIC IIT Delhi Sonipat Innovation Foundation and backing from programs like Microsoft Unnati and SIDBI [Perplexity Sonar Pro Brief][BioSky, Oct 2026]. This is classic seed-stage scaffolding: grant money and accelerator support to prove the model before a priced equity round. The next twelve months will be about converting those first four name-brand clients into referenceable, revenue-generating contracts that can attract that venture capital.
For a 100-megawatt solar farm in India, a single day of bad forecasting can trigger six-figure penalties. If BioSky’s platform can reliably shave even 20% off that risk, the value is straightforward. On the back of an envelope, for a client facing ₹1 crore in annual penalties, a 50% cut is ₹50 lakh in saved cost. That is the kind of math that makes a procurement officer listen. The company is not trying to beat the weather; it is trying to beat the generic, global forecast models that large energy traders have relied on for years. That is the incumbent it must out-perform.
Sources
- [Perplexity Sonar Pro Brief] BioSky Space Innovations Brief
- [BioSky, Oct 2026] Energy Resilience from Space | https://angel.co/jobs/signup
- [APCTT, Mar 2026] BIOSKY SPACE INNOVATIONS | https://angel.co/jobs
- [The Rocket Media, 2026] Profile of Saurabh Kapil
- [Crunchbase, Retrieved 2026] BioSky Company Profile
- [JobLeads, 2026] Founding Senior Engineer Job Listing | https://www.jobleads.com/in/job/founding-senior-engineer-power-systems-digital-intelligence--new-delhi--eea70dc41fe642adb4551a505c23260a7