Seattle-based American startup Constellation Space has unveiled its core software platform for managing satellite constellations and ground communication networks. The system integrates weather forecasts, RF interference assessments, channel loads, and space situational awareness data to calculate how much information a specific spacecraft can transmit to Earth during a single communication session. According to the company, using the platform increases the volume of transmitted data per session by up to 36%.
The Scale Problem: Thousands of Craft in Low Earth Orbit
The startup is responding to one of the key trends in the modern space industry — the rapid growth in the number of satellites in low Earth orbit. Managing constellations of thousands of craft and selecting optimal intervals for data transmission is becoming increasingly complex. For each satellite, operators must account for ground station availability, radio channel conditions, weather conditions, and potential interference. Planning errors lead to lost data transmission opportunities, which on the scale of a commercial constellation translates to direct financial losses.
Throughput Forecasting: How the Platform Works
The Constellation Space platform calculates a throughput forecast for each communication session taking a combination of factors into account. Analyzed parameters include space situational awareness data, weather, signal-to-noise ratio, radio interference, and throughput demand. Based on the predicted volume of transmitted information, operators can adjust transmission parameters and communication session schedules, minimizing downtime and maximizing the return from each satellite pass over a ground station.
Machine Learning and Adaptation to Specific Infrastructure
The machine learning model underlying the platform is pre-trained on data from a global network of sensors and, after deployment, adapts to the specific satellite constellation, its hardware, and ground stations. This means the system takes into account the infrastructure features of each operator rather than using a single set of settings for all satellites. This approach accounts for differences in frequency ranges, antenna types, geographical locations of stations, and specific orbital parameters.
The Team: From NASA and Starlink to Blue Origin
The company was founded in 2025 by Kamran Majid, who previously completed an engineering internship at NASA and worked as a software engineer on SpaceX’s Starlink satellite network team. According to him, his experience at Starlink showed that manual mission management processes can no longer cope with the growth of satellite constellations. Co-founders include Chief Technology Officer Raed Kabir, formerly of Blue Origin, as well as former Capital One engineers Omid Tehrani and Lait Altarabishi. At the time of the platform's introduction, the team consists of six people.
Investment and First Customers
Despite its young age, the startup has raised $7.5 million in a seed investment round. Financing participants include accelerator Y Combinator, Nvidia, OpenAI, and Samsung. According to Majid, an early version of the platform has already attracted more than 10 customers, including Amazon Web Services. It is also used by more than 20 academic institutions. The company's annual contract value reached $2.25 million this year, indicating rapid commercial acceleration for a startup founded less than two years ago.
