Ouster - The eyes of AI

Introduction

There is this funny thing that happens with hardware companies when they start threading software through their products. The market tends to keep valuing them as hardware companies right up until the moment it does not.

In some way, Ouster reminds me of an early Axon in term of business trajectory:

• Uniquely positioned in its respective segment
• Started as a hardware business
• Threaded software into their stack
• Created a strong interlocking flywheel, resulting in increased customer value, expanding margins and a stock that followed suit

OUST is positioning itself as the physical AI sensing and perception platform, beyond simply being a hardware supplier.

The business recently popped up on my radar after Daniel Koss (@daniel_koss) and PPDD (@usppdd) posted about it on X. What pushed me look into the business deeper, was a combination of things that I love to see:

• Perfectly positioned to benefit from several major durable tech trends:
  • The rise of psychical AI
  • Auton0mous driving
  • Drones
  • Smart infrastructure
  • Robotics
• Founder led with a highly driven and motivated management team
• A strong and innovative company culture
• 12 consecutive quarters of strong revenue growth, gross margins expanding from the high teens to nearly 50% and a strong balance sheet

After this write-up, you'll have a good understanding of what Ouster does, who runs the business, how they are positioned within their industry, and how all these pieces synthesize into an investment thesis.

1 - Management

Angus Pacala co-founded Ouster in 2015 and has run it since day one. He hold a BS and MS in mechanical engineering from Stanford, and before Ouster he was Director of Engineering at Quanergy, one of the original solid-state LiDAR companies. I personally always like leaders who have had their fingers in the dirt, so to speak. They know the space and what works.

Mark Frichtl is co-founder and CTO and the technical vision behind Ouster. This kind of founder-level technical leadership is what I love to see. Mark holds a Bachelor of Science degree in engineering physics and a Master of Science degree in mechanical engineering from Stanford University.

On the CFO side, Kenneth Gianella joined in April 2025, bringing over 20 years of financial and operational leadership experience. Prior, he was CFO at Quantum Corporation. His track record in scaling global organizations is a positive signal.

In April 2026, Ouster promoted Cyrille Jacquemet to Chief Revenue Officer. Jacquemet has been at the company since 2018 and was Senior Vice President of Global Sales since 2023 and with this new role OUST is positioning itself for what Pacala has called its "next growth chapter."

In short

This is a founder-led business with deep technical roots. The recent CFO transition introduces some execution uncertainty, but the core leadership team is long standing and has a proved track record.

2 - Industry context

LiDAR, short for Light Detection and Ranging, is the technology that lets machines see the physical world in three dimensions with precision that cameras and radar cannot match alone.

Where a camera captures light intensity and color, and radar gives coarse distance data, LiDAR emits laser pulses and measures the time they take to return, building a dense point cloud of the environment in real time.

The global LiDAR market was worth roughly $3 billion in 2025 and is forecast to reach $14 billion by 2035, growing at roughly 18% per year. The solid-state segment, which is where Ouster's Chronos chip sit, is growing much faster, with projections of $3.2 billion by 2030 from $500 million in 2025, a 45% CAGR.

But that's just the hardware part. Where I think it gets extra interesting is the software part of it.

For most of LiDAR's commercial history, the value was in the hardware. You sold a sensor, the customer integrated it, and that was that. What is changing is that the real bottleneck for autonomous system deployments is no longer the sensor. It is perception: the ability to turn raw data into actionable intelligence in real time.

"When you think about us and you want to invest in us, number one is the AI software solutions. It's not just about the hardware, it's the platform." ~ Pacala, August 2025,

The companies that own the software layer will be structurally stickier and higher-margin than pure sensor vendors. It is however good to note that software is not yet a major part of their revenue, but it's growing significantly.

"Software-attached bookings more than doubled in 2025 and represented over 15% of our sensors shipped, which is up over 120% year on year."

This is where I see close similarities with Axon. They used to be a hardware business but as of today, their software platform is their core growth driver and nearly 50% of revenue. I can see a similar path for Ouster.

3 - Mission and the problem

Ouster mission is to build a safer and more sustainable future through high-resolution digital lidar sensors for the automotive, industrial, smart infrastructure, and robotics industries.

The core customer pain is not just about sensors. It is about deployment: autonomous systems can only be deployed commercially when the full perception stack, from sensor to software to output, is reliable, certifiable, and cost-effective enough to justify the capital outlay.

The company serves four end markets: automotive, industrial, robotics, and smart infrastructure. Each one has a different adoption curve and a different margin profile. Smart infrastructure and industrial are further along in real deployments today. Automotive is the largest eventual TAM but requires the longest qualification cycles.

Where Ouster sits in the value chain has been shifting. It started as a sensor manufacturer. With Gemini (its real-time perception engine for object detection and tracking) and Blue City (its data analytics platform for traffic operations and smart infrastructure), it moved toward the platform layer.

They also acquired StereoLabs which adds stereo cameras, AI compute, sensor fusion, and pre-trained AI perception models. The company is now positioning itself as a single-source supplier for the full sensing and perception stack. That is a very different business model from where it started.

What I find particularly interesting with Ouster is that they are exposed to several major tech trends ranging from the robotics buildout, humanoid platforms, smart city investment cycles, autonomous mining and logistics applications.

4 - The technology explained

Traditional LiDAR systems are built from hundreds of discrete analog components: individual emitters, detectors, and signal circuits wired together at high cost and low scalability.

Think of it like an old telephone switchboard. Ouster replaced that entire switchboard with two chips. A custom CMOS System-on-a-Chip with integrated Single Photon Avalanche Diode detectors, and a Vertical Cavity Surface Emitting Laser array (VCSEL). Fun fact: this is the same laser technology in iPhone Face ID.

Building on CMOS silicon means Ouster sits on Moore's Law: each chip generation improves performance at lower unit cost, the same way smartphone cameras compound without the underlying physics changing. The L3 chip powered all REV7 sensors. The Chronos chip is a parallel track designed for solid-state Digital Flash applications targeting automotive and industrial OEMs.

REV8 and the L4 Silicon Generation. On May 4, 2026, Ouster announced the REV8 family powered by L4 Ouster Silicon, delivering double the range and double the resolution of REV7. That some serious technological progression. It's the first commercially available native color LiDAR sensor: color and 3D depth merged through physics at the point of capture.

The family spans four models:

• The OS1 MAX, 200m range, automotive-grade
• OS1, mid-range, smart infrastructure and AV
• OS0, 90-degree field of view for robotics and warehouses
• OSDome, 180-degree hemispherical for full-surround coverage

The StereoLabs acquisition layered ZED stereo cameras, AI compute, and pre-trained perception models on top of this hardware foundation, across more than 90,000 deployed cameras and 10,000 customers.

The combination of REV8 native color LiDAR and ZED creates what no competitor currently offers at scale: spatially aligned color-depth sensing with AI perception built in. You cannot replicate this stack quickly. Silicon design cycles run two to three years per generation. Safety certifications take longer and customer integrations are sticky from the moment a design is qualified.

Written by

Share

Copied