From Action Cams to AI Photographers: What Cloud Migration Can Learn from Insta360

The Camera That Forgets Itself

Ten years ago, a 360-degree camera was a novelty that needed explaining. Insta360's first model, the Nano, plugged directly into an iPhone to cut down transfer friction. It was an accessory, a toy for early adopters. The workflow was painful: shoot, transfer, stitch, then scrub through a full sphere to find the one angle you actually wanted.

Today, the company's X-series has evolved far beyond that. Over half of Insta360's customers now buy these cameras not for extreme sports, but for travel, family gatherings, graduations, and weddings. The product has quietly shifted from a niche action tool to a mainstream memory-capture device.

But founder Liu Jingkang has a counterintuitive vision. After a decade of building cameras, he wants to make one that people forget they're wearing. He calls it 'unconscious' or 'senseless' recording. The goal isn't a better camera—it's a camera that disappears into the moment.

The Hidden Cost of 'Shoot Now, Frame Later'

The original pitch for 360-degree video was liberating: you don't have to compose a shot before pressing record. The entire space is captured, so you can decide what to show after the fact. That's why early users strapped the Nano to helmets and selfie sticks for skiing and cycling—when you're moving fast, you don't have the attention span for framing.

But this freedom created a new burden. After the shoot, someone has to edit. And editing 360-degree footage is miserable. You have to watch the whole sphere, pick the moment, choose the perspective, and then cut it into a watchable clip. For most people, that's too much work.

Liu Jingkang believes this is the core problem the industry hasn't solved. 'Our mission is that the ideal camera should make you forget it exists,' he says. 'So we introduced zero-editing—the camera itself, while charging and syncing to the cloud, automatically selects the best shots for you.'

Cloud Migration: The Invisible Backbone

This is where cloud migration enters the picture. The X6 doesn't just rely on on-device AI. It uses a hybrid approach: edge processing for real-time capture and basic analysis, plus cloud-based models for deeper understanding and automated editing. When you plug the camera in at night, it uploads footage to the cloud, where a multimodal AI model—trained on over 10,000 hours of travel, cycling, diving, and skiing footage—identifies highlights, designs virtual camera moves, and assembles a finished video by morning.

This is a classic cloud migration pattern. Instead of trying to do everything on a tiny device with limited battery and compute, you offload the heavy lifting to scalable cloud infrastructure. The camera becomes a thin client for capturing raw data; the cloud becomes the brain that turns it into a story.

For enterprises, the analogy is direct: you don't run massive analytics workloads on a laptop. You migrate to the cloud, where you have elastic compute, storage, and AI services. Insta360 is doing the same for personal video—and it's a lesson in knowing what to keep at the edge and what to move to the cloud.

From Edge to Cloud and Back Again

But the migration isn't one-way. Liu Jingkang points out that today's X6 edits while charging because on-device AI and battery capacity are still limited. The cloud offers more power, but it also introduces latency and privacy concerns. As chips get faster and more efficient, he sees a future where the camera can edit instantly on the device—'shoot and edit in real time.'

This mirrors the hybrid cloud trend in enterprise IT. You keep latency-sensitive workloads at the edge, push batch processing to the cloud, and continuously rebalance as hardware improves. The optimal split changes every few years. Insta360 is already thinking about that dance.

They're also exploring cloud-based 3D reconstruction, using Gaussian splatting to let users move through a captured space—not just change the viewing angle, but change the viewpoint itself. That kind of heavy computation will likely live in the cloud for a long time, until edge hardware catches up.

The Market Is the Problem, Not the Cake

When asked about the market size for 360-degree cameras, Liu Jingkang dismisses the question. 'Market size isn't predicted by a model,' he says. 'It depends on how many problems you solve.' He compares it to bottled water and tissue—products that became ubiquitous because they solved universal needs.

'There are still two orders of magnitude between 360 cameras and smartphones,' he notes. 'That's a lot of unsolved problems. And problems are markets.'

This is a useful lens for cloud migration. Instead of asking, 'How big is the cloud migration market?' ask, 'What problems are still unsolved?' For many companies, the answer is: data silos, security compliance, latency, cost unpredictability, and the shortage of skilled cloud architects. Each of those is an opportunity—not just to lift and shift, but to redesign workflows around cloud-native capabilities.

Recording, Not Creating

Liu Jingkang draws a sharp line between 'creation' and 'recording.' Creation is an intentional act—choosing a subject, composing a shot, editing for effect. That market is already well-served by professional cameras and editing software. Recording, on the other hand, is about capturing moments without distraction. 'You can't hire a photographer for every precious moment,' he says.

This philosophy drives Insta360's product decisions. The X6 is smaller, has longer battery life, and automates editing because the target user isn't a filmmaker. It's a parent at a kid's soccer game, a traveler on a crowded street, a couple at their wedding. They want the memory, not the craft.

For cloud migration, the lesson is similar: don't force users to become IT experts. The best cloud solutions are those that abstract away complexity—serverless functions, managed databases, and AI services that let developers focus on logic, not infrastructure. The cloud should be the invisible enabler, not the thing you have to think about.

The Future: AI Photographers and Spatial Records

Looking ahead, Liu Jingkang envisions 'AI photographers' that understand what you care about. The head-tracking module on the X6 already lets you 'look where you want to shoot,' and the AI learns your editing preferences over time. Eventually, he predicts, the camera will know what to capture and how to cut it based on your past behavior—a 'thousand-person, thousand-face' approach to video.

Beyond that, he sees the product evolving from capturing video to capturing space. The 3D reconstruction demo he shows allows you to walk through a recorded environment, not just rotate your view. That's a fundamentally different kind of record—a spatial memory, not just a flat image.

This has implications for cloud migration, too. As we move from storing files to storing immersive experiences, the cloud will need to handle massive amounts of 3D data, real-time rendering, and AI-driven scene understanding. The migration isn't just about moving current workloads—it's about preparing for workloads that don't exist yet.

What Cloud Migrators Can Steal from Insta360

Insta360's journey offers three takeaways for anyone planning a cloud migration:

• Start with the user's pain, not the technology. Insta360 didn't build a better camera for the sake of specs; they solved a specific problem—the drudgery of editing 360 footage. Similarly, migrate to the cloud to solve a business pain, whether it's cost, agility, or scale.
• Hybrid is a strategy, not a compromise. The X6 uses edge for capture and cloud for heavy AI. Find the right split for your workloads. Some things belong at the edge; others belong in the cloud. Revisit the split as technology improves.
• Automate the boring parts. Insta360's killer feature is zero-editing. The cloud equivalent is automation—infrastructure as code, auto-scaling, managed services that reduce toil. The less your team has to think about plumbing, the more they can focus on your actual product.

The Cloud Is the New Camera

Liu Jingkang's ultimate vision is a 'photography robot'—a device that autonomously finds the best angle, captures the moment, and edits the story, all without human intervention. He's even open to the possibility that the physical camera might disappear, replaced by software that reconstructs spaces from ordinary photos.

For cloud migration, the parallel is clear: the cloud is becoming the invisible processor of all our digital experiences. Just as Insta360 wants you to forget the camera, the best cloud architectures are the ones you don't notice—they just work, scale, and adapt. The migration isn't a one-time project; it's a continuous evolution toward a state where the infrastructure is as unobtrusive as a good assistant.

Ten years ago, Insta360's name meant 'instant capture.' Today, it means 'instant story.' For cloud migrators, the lesson is to aim for that same shift—from simply moving data to creating a seamless, automated, and ultimately invisible infrastructure that powers your business. The camera is just the beginning.