Founding Machine Learning Scientist

About BioSky

BioSky is a deep-tech spacetech startup building physics-driven, satellite-AI solutions for energy, climate, and sustainability use-cases. We work at the intersection of geospatial intelligence, large-scale ML, and real-world deployment , turning multi-sensor satellite data into business-critical insights.

We’re funded and backed by IIT Delhi , and selected among the Top 15 global spacetech startups by:

• Venture Catalyst Space (Australia)
• AWS Space Accelerator
• TACC+ Taiwan International Spacetech 2025

The Role

We’re looking for a Founding Machine-Learning Scientist who has already built and deployed ML models in production and now wants to own the entire product ML stack —from raw satellite data to real-time APIs.

This is a zero-to-one role . You’ll work on unstructured, high-impact problems and ship fast.

What You’ll Do

• Design, train, and deploy deep-learning models (CNNs, LSTM/RNNs, Transformers, GNNs) on satellite & geospatial data
• Build end-to-end ML pipelines : data ingestion → feature engineering → training → deployment
• Work on multi-sensor data fusion (satellite + external data sources) at scale
• Own the complete ML lifecycle : experimentation, evaluation, MLOps, CI/CD, cloud inference
• Compress petabyte-scale data into millisecond-latency APIs
• Rapidly iterate models based on real product feedback—not just benchmarks

Who You Are

We are specifically looking for engineers who:

• Have 2–3 years of hands-on experience deploying ML models in startups or fast-moving teams
• Enjoy working with messy, unstructured, real-world data
• Think like product builders , not just model trainers
• Are comfortable with ambiguity and take full ownership
• Want to experience the early-stage startup journey deeply
• Have a long-term goal of starting their own startup and want to learn by building

Must-Have Skills

• Strong fundamentals in ML, statistics, probability
• Experience with time-series and image models (ConvLSTM, Attention, Vision Transformers)
• Proficiency in Python (PyTorch / TensorFlow, scikit-learn)
• Experience building production ML systems , not just notebooks
• Bachelor’s or Master’s in CS, Data Science, or related fields

Bonus (Nice to Have)

• Experience in geospatial ML / remote sensing / satellite imagery
• Industry-scale time-series forecasting or computer vision
• Hands-on MLOps , automation, and cloud GPU training (AWS/GCP)

What You Get

• Competitive salary + Founding-team ESOPs
• End-to-end ownership of ML features and products
• Opportunity to build real-time spacetech products
• Mentorship from experts at ISRO, IIT Delhi/Bombay, Australian Space Agency, SmartSat CRC
• Flat structure, fast experimentation, strong tech infrastructure

Team

Founded by alumni of IIT Bombay and IIM Calcutta , with prior experience at ISRO, Strategy&, PwC , and global consulting & research organizations.