Perceptive Space

Perceptive Space is a Canadian AI-driven space weather prediction startup founded in 2022 and headquartered in Toronto, Ontario, developing machine learning-powered forecasting systems for space weather events — including solar flares, geomagnetic storms, and ionospheric disturbances — that the company claims outperform traditional physics-based models by approximately 10x in accuracy, forecast speed, and operational reliability.

The company was founded by Padmashri Suresh, a former NASA-sponsored space weather researcher who combined domain expertise in solar-terrestrial physics with AI/ML modeling techniques. Traditional space weather forecasting relies on physics-based magnetohydrodynamic (MHD) simulation models that require significant computational resources and have inherent lead-time constraints — typically providing meaningful geomagnetic storm warning of only 15-60 minutes from real-time solar wind monitoring at the L1 Lagrange point by spacecraft such as DSCOVR and ACE.

Perceptive Space trains large machine learning models on decades of historical space weather observational datasets — solar imagery, magnetometer data, particle flux measurements, ionospheric data — to identify precursor patterns and signatures enabling earlier and more accurate forecasts. Extended warning times give satellite operators, power grid operators in auroral regions vulnerable to geomagnetically induced currents (GICs), and aviation customers more time to take protective measures before the onset of severe space weather events that can damage satellite electronics, disrupt GPS accuracy, and cause grid failures.

The company raised CAD $3.9 million (approximately $2.9M USD) in pre-seed funding to develop and commercialize its forecasting platform, competing with NOAA's Space Weather Prediction Center and commercial providers like Metatech and SpaceAble in the growing commercial space weather services market.