In our ongoing Meet the Member blog series, we talk to members of the Satcoms Innovation Group (SIG) to find out more about their offering and to see how they are working to resolve the operational and technical challenges that satcom currently faces.
In the latest blog post in the series, we meet Serana AI who recently joined SIG as a startup member.
Serana AI’s profile
Serana AI, a Canadian startup, was founded by Najeeb Khan (PhD in machine learning, previously at Amazon and Calian’s space and defence business), with advisors from the University of British Columbia and the Canadian space sector.
The company is building an agentic planning platform for satellite operators. With the platform, an operator will be able to ask a planning question in plain language, for example whether a beam can serve a new customer at the required throughput once rain fade is accounted for, and Serana works through the relevant physics and live fleet state to answer in seconds, with every step laid out for an engineer to verify.
Currently, answering that kind of question can take days, spread across spreadsheets and disconnected tools that each handle one constellation, beam, or link. Real decisions span demand, weather, spectrum, routing, SLA risk, and interference at once.
The destination is autonomous connectivity: networks that plan, optimise, and ultimately heal themselves.
What industry challenges are having the biggest influence on your work at the moment?
Planning complexity is outpacing the tools the industry was built on. Multi-orbit fleets, denser constellations, and increasingly dynamic demand have multiplied the variables behind every capacity and link-budget decision. Interference is now one of those variables, not a separate problem to be handled later.
Legacy, siloed tooling was never designed for this, and the experienced engineers who can hold it all together are scarce and in high demand. Operators need faster, more reliable and precise answers as the questions become harder.
What innovative solutions are you working on to address the issue?
The hard part of automation in this industry is not generating an answer, it is trusting one. Our agents reason over real physics and live fleet data rather than text alone, so recommendations are grounded in the operator’s actual network. Crucially, our “Why Not?” method produces solver-grounded certificates: every recommendation arrives with a verifiable explanation an engineer can audit, not a black box. That auditability is what makes automation credible in mission-critical operations, and it is the basis of a provisional patent we have filed.
What are your thoughts on spectrum management?
Spectrum is fast becoming the binding constraint on industry growth, and rising interference, the issue this group was originally formed to tackle, is the clearest symptom. As LEO, MEO, and GEO systems increasingly share the same bands, static frequency plans and manual coordination cannot keep pace.
We believe the path forward is software-defined, interference-aware planning: tools that model and reason about RFI dynamically, helping operators protect service quality and coordinate proactively rather than react after the fact. It is an area where automation can deliver real operational and commercial value, which is exactly why it sits at the centre of our roadmap.
Do you have any interesting plans for the next couple of months?
Over the next few months, we are moving our first operator engagement from demo toward a working deployment, showing that agentic planning can support real operational workflows across both GEO and NGSO. We are also developing predictive interference modelling, shifting from detecting RFI to anticipating it before it affects service. Underpinning the platform is a foundational LLM we are building to understand the satcom domain natively, grounded in real physics, regulation, and operations.
Why do you feel it is important to be a member of SIG?
SIG sits exactly where our work lives: operators, engineers, and innovators tackling shared challenges, interference foremost among them. For a company building decision tools for this community, there is no substitute for proximity to the people who run these networks and the technical and standards conversations that shape what is possible. SIG keeps us close to real operator priorities and lets us contribute to problems the whole industry has a stake in solving.
To find out more about Serana AI, visit https://serana.ai
To find out more about SIG and to become a member, visit https://satig.space