When we inherited a React Native app, users were constantly running into a critical problem. They’d record a video, hit submit, then close the app or switch away. Everything looked fine on the surface, but in reality, the upload would fail silently. There was no retry prompt. No progress bar. No confirmation. Videos were simply lost.
From the user's point of view, they followed the flow correctly. From the system's point of view, the submission was never completed. We knew this wasn’t just a bad UX, it was a serious breakdown in trust and functionality.
So we rebuilt the entire video upload pipeline to be fast, resilient, and transparent.
What we changed
1. Efficient Video Compression with Custom Native Modules
We evaluated several libraries for client-side video compression but found none that were reliable, maintained, and integrated cleanly into our stack. So we created custom native modules for both iOS and Android, using platform APIs to compress the video prior to upload.
This significantly reduced file sizes and improved the success rate of uploads, especially for users on slower networks. We avoided unnecessary transcoding steps and preserved enough quality for the app’s requirements.
Compression was critical, but it had to be done in a way that felt native, fast, and invisible to the user.
2. Chunked Uploads with Recovery Logic
Once videos were compressed, we split them into chunks for upload. If a chunk failed, we only retried that specific part instead of restarting the entire transfer.
On the backend, we accepted and tracked incoming chunks, storing them temporarily until all expected parts were received. Once complete, we stitched them back together and finalized the submission.
This gave us the flexibility to recover from partial failures and created a more durable upload process for users moving between Wi-Fi and cellular or in low-signal areas.
3. Backend-Handled Background Upload Continuation
Rather than forcing the mobile app to keep uploads alive when backgrounded, we shifted responsibility to the backend.
Once the frontend submitted the compressed video and initial metadata, the backend handled the rest, even if the app was closed. This gave us far more control over retries, completion logic, and monitoring.
By offloading responsibility to the backend, we were able to create a reliable, event-driven pipeline that continued processing uploads regardless of app state.
This eliminated one of the largest points of failure and improved both reliability and user confidence.
4. API Call Optimization for Performance
While improving the upload flow, we also revisited how the app interacted with the backend during submission.
We noticed several redundant or sequential calls in the original flow. By combining related backend endpoints and batching multiple actions into a single request where appropriate, we reduced the number of API calls required to complete a submission.
This led to faster perceived performance, lower network overhead, and a more responsive UI during and after upload.
5. Clear UI Feedback and Retry Options
We added progress indicators, confirmation states, and error handling to make the process transparent.
If a user’s upload failed or stalled, they would now see exactly where the issue occurred and have an option to retry immediately. No more silent failures. No more guessing.
This made the app feel responsive and gave users control over what happened next.
The result
Upload failures dropped dramatically. Video submissions became faster, more resilient, and easier to trust. Users no longer had to keep the app open or babysit the process. They could simply hit submit and move on, knowing it would be handled.
On the technical side, the backend now manages heavy lifting. Compression is fast and reliable. Uploads are chunked and recoverable. API usage is more efficient. And the frontend experience finally matches user expectations.
If your app involves media uploads, don’t underestimate the impact of background processing, server-side responsibility, and smart compression.
When users submit content, they’re trusting your system to handle it. That process needs to be rock solid, even in the messiest network conditions.
We’ve rebuilt this flow from the ground up and seen the results firsthand. If your team is facing similar reliability issues or you’re planning to implement video upload at scale, we’d be happy to help.
Let’s make your app something users can count on.
.png)