Building Traffic Lights for Space: Lilian Krengel on AI, Orbital Congestion, and Student Founding

Lilian Krengel is nineteen. She's also the founder of OrbitGuard AI, a startup tackling something most people can't see and don't yet worry about: low Earth orbit is filling up. Starlink is sending hundreds of satellites a week. Old satellites can't easily come down. Physicists have a name for what happens when one collision triggers a cascade. They call it the Kessler Effect.

In this Still Human conversation, Lilian explains why she didn't try to build a rocket. She built the traffic management layer instead. Air traffic control, but for space. Her AI takes raw conjunction data and turns it into a clear collision risk score that satellite operators can actually act on, integrated with KeepTrack — the orbital visualization tool the industry already uses.

Traffic Lights for Space

Air traffic control exists because thousands of planes share the same sky and nobody trusts pilots to negotiate it by hand. The same thing is now true above us — except the planes are satellites, the sky is low Earth orbit, and the negotiation layer doesn't exist yet. OrbitGuard's job is to be that layer: ingest conjunction data, output a clean risk score, integrate with the visualization the industry already uses. Routing, not rockets.

The Kessler Effect, Explained

The textbook version: one satellite collision creates debris, that debris hits more satellites, the cascade saturates the orbit, and eventually no one can launch into it at all. The reason it matters now is launch cadence. Starlink alone is putting hundreds of satellites up per week. Each one is a future piece of debris if it can't deorbit cleanly. The Kessler Effect is what happens if traffic management arrives a decade too late.

Why Infrastructure Beats Hardware

Most aerospace founders want to build a rocket. Lilian deliberately didn't. The strategic bet is that infrastructure — the layer that everyone has to use regardless of which satellites end up in orbit — compounds differently from a single piece of hardware. A rocket competes with other rockets. The traffic management layer doesn't compete with anyone yet, and once it's adopted, it becomes the default.

Getting Copied in a Weekend

A Stanford hackathon team built something almost identical to OrbitGuard in a weekend. Most founders would treat that as catastrophic. Lilian's response was a recalibration — not "we're dead" but "this is what AI-era execution speed looks like, and we have to move faster." The lesson generalizes: if your idea can be cloned in 48 hours, the moat isn't the idea. It's the people who keep showing up to extend it.

AI in Education, Done Right

Schools banning AI from classrooms are losing the bigger argument. Lilian's framing: teach kids to code before they prompt, so AI amplifies real understanding instead of replacing it. The fear that AI will erase student effort is real — but the fix isn't a ban, it's sequencing. First the fundamentals, then the multiplier. The kids who never built understanding get hollowed-out output. The kids who built it first get superpowers.

The Capacity Question

Lilian runs a startup, sits on student senate, manages a 500-person club's finances, and is carrying a 12.5-hour sleep debt. Capacity is a skill. It's not aspirational; it's the unsexy actual cost of building something while you're still in school. The honest version of student founding isn't a montage. It's deciding which hour of sleep you owe back to which week, and showing up to class anyway.

Show Notes

Lilian Krengel is a sophomore at Santa Clara University and the founder of OrbitGuard AI, a real-time satellite collision risk system that ingests conjunction data and outputs a clean risk score human operators can act on, integrated with KeepTrack. She came up through 13 years of competitive gymnastics, argued NATO AI policy in collegiate-format policy debate at 16, and made an early strategic decision most aerospace founders don't: build infrastructure, not hardware. Air traffic control, but for space. For the Still Human audience, Lilian is the guest who reframes the AI economy as a thing happening 400 kilometers above your head — and shows what student founding looks like when the problem is real and the timeline is genuinely urgent.

Articles & Research

No external research was cited in this episode.

Tools & Resources

Relevant to this episode:

• OrbitGuard AI — Lilian's startup; ingests raw conjunction data and outputs a collision risk score for satellite operators
• KeepTrack — The orbital visualization tool already used by the industry; OrbitGuard integrates with it
• The Kessler Effect — The cascade scenario in low Earth orbit where one collision creates debris that triggers more collisions; the underlying threat OrbitGuard exists to mitigate
• Starlink and the NewSpace launch cadence — Hundreds of new satellites per week; the context for why orbital traffic management is suddenly urgent
• Policy debate (collegiate format) — Where Lilian first argued both sides of NATO AI governance at 16; referenced as formative training for systems thinking

Related Still Human Episodes

You might also enjoy:

• AI Is the Ultimate People Pleaser — Bailley Georgieva on Hypersonics & Critical Thinking — oshenstudio.com/episode/ai-people-pleaser-bailley-georgieva-hypersonics
• Execution Culture — Sean Wu on Raising $2M for Robotics & the Sim-to-Real Gap — oshenstudio.com/episode/execution-culture-sean-wu-synphony-robotics
• Build Before You're Ready — Andrey Marey on High-Agency, Discipline & Refusing AI With Friends — oshenstudio.com/episode/high-agency-andrey-marey-student-founder

People Mentioned

• Bailley Georgieva — Still Human episode 4 guest; her question to Lilian — "who are you without AI?" — appears around the 44:00 mark

Key Takeaways

• Build the traffic lights, not the rocket. Lilian's strategic bet is infrastructure over hardware — the layer that has to exist regardless of which satellites end up in orbit.
• The Kessler Effect is a real risk model, not a sci-fi trope. Once orbital debris reaches a certain density, collisions cascade. OrbitGuard exists because the timeline matters.
• Industry adoption beats clever from scratch. Integrating with KeepTrack — what operators already use — makes the risk score useful on day one instead of in year three.
• Execution speed is the moat in the AI era. A Stanford weekend project built something nearly identical to OrbitGuard. Lilian's response wasn't despair; it was a recalibration of how fast student founders now have to move.
• Banning AI in schools is the wrong fight. Lilian's case: teach kids to code before they prompt, so AI amplifies real understanding instead of replacing it.
• Capacity is a skill. Founder + senate + 500-person club finances + classes + a 12.5-hour sleep debt. Not aspirational — the actual cost of building something while you're still in school.

In This Episode

• What "traffic lights for space" actually means — Lilian's framing: air traffic control, but for low Earth orbit, with AI doing the routing
• The Kessler Effect in plain language — Why one collision can trigger a cascade, and why this stops being abstract once Starlink is launching at the current cadence
• How OrbitGuard works — Conjunction data in, collision risk score out, integrated with KeepTrack so operators can act inside their existing workflow
• Why she chose infrastructure over hardware — The strategic bet behind not trying to build a rocket
• Getting copied in a weekend — The Stanford hackathon moment, what it actually felt like, and what it taught her about execution speed in the AI era
• AI in education — Lilian's case for teaching code before prompts, and why she thinks schools that ban AI are losing the bigger argument
• The capacity question — Running a startup while sitting on student senate, managing a 500-person club's finances, and carrying a real sleep debt — the unsexy cost of building during undergrad
• Who are you without AI? — Lilian's answer to Bailley Georgieva's question, passed forward from episode 4

About Lilian Krengel

Lilian Krengel is a sophomore at Santa Clara University and the founder of OrbitGuard AI, a startup focused on space traffic management — turning raw conjunction data into actionable collision risk scores for satellite operators, integrated with KeepTrack. Her path runs through 13 years of competitive gymnastics, collegiate-format policy debate where she argued both sides of NATO AI governance at 16, and a deliberate strategic decision early on: to build the orbital infrastructure layer rather than another rocket. Lilian's view on AI is grounded — treat it like a calculator, a powerful extension of human judgment that doesn't substitute for the thinking underneath. For the Still Human audience, she's the guest who reframes the AI economy as something happening above your head, on a timeline that's genuinely urgent, and shows what serious student founding looks like when the problem is real.

Connect With Lilian Krengel

• Watch the episode: youtube.com/watch?v=I3lR2WbpVy8
• Add Lilian's LinkedIn, OrbitGuard website, or other socials here

Follow Still Human Podcast

Still Human Podcast is a biweekly show by Oshen Studio, hosted by Perkin — exploring what it means to stay human in the age of AI. Real conversations with builders, creators, founders, and thinkers doing it in real life.

• YouTube: youtube.com/@Oshen.studio
• Instagram: instagram.com/perkin0909
• Listen everywhere: Search Still Human Podcast on Spotify, Apple Podcasts, or wherever you listen

New episodes drop every two weeks. Subscribe so you never miss a conversation.