Real Alpha
Watch modeSubscribed

The Shawn Ryan Show · 3 min · Oct 23, 2025

Shield AI can integrate an AI pilot with three engineers in three months

Brandon Tseng says a reusable AI pilot now takes three engineers and three months to integrate on a Lakota variant, compared with 40 engineers and two years before. That makes it faster to bring autonomy to different platforms.

Brandon Tseng

Brandon Tseng is president and cofounder of Shield AI.

Co-founded Shield AI, maker of autonomous V-BAT drones and AI pilots flying F-16s; former Navy SEAL.

About this segment

Brandon Tseng says a reusable AI pilot now takes three engineers and three months to integrate on a Lakota variant, compared with 40 engineers and two years before. That makes it faster to bring autonomy to different platforms.

From “Brandon Tseng – Shield AI’s X-BAT: The First AI Fighter Jet to Outsmart Top Gun | SRS #247”, Oct 23, 2025. Watch the full episode

“we can now do, with a couple hundred thousand dollars, three engineers in three months”
Brandon Tsengat 1:42:58 · The Shawn Ryan Show1:42:58

Takeaways

  1. Earlier quadcopter integration took 40 engineers and about two years.
  2. A Lakota variant took three engineers and three months.
  3. Shield AI says its AI pilot has been put on 15 platforms.

In this episode

Full episode
  1. 1:51:22X-BAT aims to deliver fighter capability without runways7:14
  2. 1:15:48Autonomous systems learn through simulation and reward5:02

  3. 1:40:38Shield AI can integrate an AI pilot with three engineers in three months2:54

  4. 1:13:20Shield AI moved beyond a small quadcopter market2:51Up next
  5. 6:15V-BAT supported a major Caribbean drug interdiction2:23

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

  1. 1:40:51

    yes we do yeah was it always like that um it was it was out of a necessity right when we wanted uh And we started off with the quadcopter, we said, hey, are there any other quadcopters that can carry the sensors, the compute, and the package that we need to do to make a product? And the answer was like, no, those didn't exist. And so we decided to, hey, we're gonna have to build our own quadcopter. So we built a hardware engineering team, built that product.

  2. 1:41:18

    and then when we decided to like move up the aviation food chain um a big reason for that was what like i could tell someone i could tell a four-star general four-star admiral or like uh you know an under secretary all day every day that this ai pilot that's flying your quadcopter you can put it on f-16 but back in like 2018 2019 everybody's like nah like i don't believe that right um so we were like hey we need another piece of hardware to demonstrate that this AI pilot is transferable from platform to platform to platform.

  3. 1:41:56

    That led to the acquisition of the VBAT. We're like, okay, now we can put the AI pilot on the VBAT. Now you see the AI pilot running on a quadcaptor. You see it running on a VBAT. We bought Heron Systems and, you know, that won the DARPA-alpha dogfight. And then we're the first company in the world to fly the F-16 completely autonomously. Now you see the AI pilot running on an F-16. And since then, now we've put the AI pilot on 15 different platforms, one way attack drones, unmanned ground vehicles, you know, large target drones, jet powered drones, where I think everybody at this stage in the game in 2025, people realize, hey, the software is actually transferable, portable from system to system.

  4. 1:42:39

    And again, this is something that we architected our autonomy stack from the get go, right? With that vision of a world of autonomous systems. And so like Shield AI stands alone in our ability to integrate onto platform, platform, platform, platform really rapidly. Like, one of the cool things, we can take... We've integrated on a variant of the Lakota helicopter. What used to take 40 engineers and, like, tens of millions of dollars and two years to do on our quadcopter, we can now do, with a couple hundred thousand dollars, three engineers in three months, on something like a Lakota helicopter or something like, you know, a jet-powered drone.