One aircraft. Every mission.
Synapse Aerospace builds autonomous aircraft as flying data centers — one platform, an interchangeable payload, and a ground station that reconfigures its own mission planning around whatever that payload is built to do.
The position
Most drones are built for one job. Ours are built for whatever job comes next.
We build the airframe once. What turns it into a mapping platform, a survey tool, or a search-and-rescue asset is the payload — and the intelligence that reconfigures the mission around it. Swap what's mounted, and the ground station unlocks a different set of mission capabilities on its own. The aircraft doesn't need to be redesigned. It needs to be told what it's carrying.
We call this a flying data center: every mission, whatever the payload, exists to produce data — and every flight makes the next one smarter.
One platform, every mission
Multi-purpose by design.
A Synapse aircraft isn't built for one job. The same airframe that maps terrain today can carry a different payload tomorrow and go looking for a property boundary — or, when it matters most, for a person.
Topography. Cadastral survey. Search and rescue. These are examples of what the platform can carry, not the limits of it.
What makes this possible isn't a bigger drone. It's a stack designed so the payload decides the mission, not the airframe. Swap what's mounted, and the ground station reconfigures around it automatically — new mission planning, new autonomy behavior, new data pipeline, without redesigning the aircraft.
This is what "AI-enabled" means for us in practice: mission planning that adapts ad-hoc to whatever the aircraft is carrying, informed by data from every flight that came before it.
The stack we own
Four layers, every one of them ours.
Read bottom to top — it's the same order the aircraft does. The flight controller keeps the aircraft alive; everything above it earns authority one bounded capability at a time.
Flight Control
The sovereign layer, and the only one every mission shares regardless of payload. Attitude, safety gates, and hard real-time guarantees that don't change no matter what the aircraft is carrying or what it's been asked to do. Written from scratch — no autopilot forks, no black box we can't open at three in the morning.
Autonomy
Perception and mission behaviour, adapted to whatever payload and mission are active. Untrusted by design: every capability is requested and granted, never assumed. Flight control stays the final arbiter. Arming never leaves the human. Right now it rides in the co-pilot's seat — reading every flight, handshaking with flight control, and earning the controls one capability at a time.
Ground Control
The operator's window into the mission: live telemetry, direct control, device management — and the layer that reconfigures itself around whatever payload is mounted. Shows only what the data can actually defend.
Mission Intelligence
This is the fleet's memory: every aircraft uploads its logs the moment it lands, bound to the unit that flew it. The system reads them, works out which parameters should change, and pushes the change back down to the aircraft. Recommendations carry their reasoning. The loop closes without anyone hand-editing a config file — and it compounds across every flight the fleet has ever made.
Flight Control → Autonomy → Ground Control → Mission Intelligence
The airframe isn't the company
The airframe is disposable. Ours get redesigned, crashed, rebuilt — that's the cost of testing in the real world, not a setback. What never gets thrown away is the software running inside it: the control loops, the perception, the tuning pipeline that carries everything it learned from the last airframe straight into the next one.
That's the actual company. A frame that flies is a solved problem. Software that makes every frame it touches smarter than the one before it is the deep-tech part — and it's the part that doesn't crash.
How we actually work
Engineered like an aerospace company. Iterated like a startup.
Our own CI, our own infrastructure, code review that doesn't bend. The aircraft doesn't care about our process — so we fail fast, iterate cheap, and never guess when a five-minute flight test can tell us the truth.
One variable per flight
Change one thing, predict the outcome before takeoff, grade it after. Multi-variable flights are how programs lose the ability to know what worked.
Measured beats predicted
Every constant in the stack — throttle curves, vibration bands, control gains — was replaced by a real measurement the same day it was first estimated.
Every failure becomes a test
A false alarm, a bad recommendation, a crash — each one becomes a permanent regression test named after the incident. We don't get to make the same mistake twice.
The aircraft is patient. The ground isn't.
If we can't source a fact to a file, a log, or a measurement, we don't ship the claim.
Not small. Unlimited.
We didn't get here by adding headcount. Expertise doesn't scale that way — you go deep, or you don't. Each of us has spent a career mastering one hard discipline, and we run all of it as a data-driven, AI-enabled operation: every result gets measured, every pattern gets used, and the model informing tomorrow's decision comes from today's flight, not from a hunch.
Curiosity decides what we chase. Data decides what we keep. The thinking is the expensive part — everything after that is just execution.
Founders
Three subject-matter experts. The combination is the asset.
Sergio Virahonda
Software Architect & AI Engineer
A decade of backend engineering in high-stakes fintech — JPMorgan, Capital One — building production-grade, event-driven systems where a mistake costs real money. Brings that same discipline, plus a foundation in electronics and telecommunications engineering, to Synapse: the companion computer, the fleet backbone, and the AI orchestration that reads every flight. Software, cloud, and hardware aren't separate disciplines to him — they're one system.
Ernesto Morales
Aerospace Engineer / Business Strategy
Trained as an aerospace engineer at Zodiac Aerospace — now part of Safran — designing aircraft components in CATIA. Went on to certify Mexico's first IoT electricity meter through the national utility lab, and now runs a 46 MW energy portfolio alongside half a gigawatt of substation development. He designs every airframe Synapse builds; his sharper edge is turning hard engineering into a company with a market and a plan to grow.
Rodrigo Cangas
Test Pilot & Software Engineer
A licensed drone pilot and a software engineer before he ever became one — two disciplines most teams split across two hires. He flies every test flight himself and feeds that judgment directly back into airframe design. Years spent using drones as data-acquisition machines taught him exactly how a payload has to be built — he's the mastermind behind Synapse's battle-proof payloads.
Let's talk about what this platform can carry for you.
We've built the full stack — airframe, autonomy, ground control, and the intelligence that ties them together — because we don't accept a ceiling on what deep expertise, curiosity, and good data can do. If you're building in this space, we'd like to talk about what a multi-purpose autonomous platform could do for your program.