Why Cloud Media Workflows Are Hard at Netflix Scale

A single Netflix title generates roughly 200 terabytes of Original Camera Files (OCF), with outliers reaching 700 terabytes—before counting work-in-progress files, VFX assets, or 3D assets. Traditionally, that data lands on physical tape stock like LTO, which is cost-effective but introduces real friction: tape must be shipped and tracked physically, and media stored on tape is difficult to search, play, or share. When titles involve talent and vendors across the globe, that slowdown becomes a creative liability.

Even fully digital pipelines strain under the weight of media distribution between departments and vendors. Without automation and standardization, post-production and VFX rely on manual, labor-intensive handoffs riddled with dependencies and vulnerability to human error. As file sizes grow, these inefficiencies compound, eating into time better spent on creative decisions. Many post facilities still keep walls of portable hard drives and hand-carry media between vendors because no better alternative exists.

The move to cloud workflows introduces its own operational hurdles. Netflix believes great stories can come from anywhere, but traditional technical limitations reduce access to media and restrict access to talent. Artists need powerful workstations and real-time playback, and in some markets that technology isn't available or affordable. The industry shift, visible at events like IBC and NAB, points toward bringing people and applications to the media rather than shipping media to artists. That vision is not new, but executing it across hundreds of titles a year at Netflix scale is uncharted territory.

Globalization Demands Standardization

Building global technology for filmmaking means serving markets at wildly different stages of maturity. Some regions have over a century of cinematic history; others are just establishing themselves. Each has distinct languages, workflows, and local nuances, and that diversity makes standardization difficult, especially when productions draw from a global talent pool.

Established post and VFX facilities often run on custom scripts and automation—for example, transcoding or transcription jobs triggered when files land in a hot folder, expecting specific sidecar metadata and directory structures. New workflows that disrupt those processes threaten vendor profit margins and adoption. Innovations must deliver clear, meaningful benefits to earn trust in markets with low tolerance for bugs and high demand for reliability.

Netflix's approach avoids manual configuration. Rather than having people write JSON files to map camera formats to deliverables, the platform leans on open industry standards: ACES and AMF for color pipeline management, ASC MHL for file management and verification, ASC FDL for framing interoperability, and OTIO for timeline interchange. Standards enable automation at scale and open high-complexity workflows to productions that wouldn't normally have access. A show shot across multiple camera formats, resolutions, lenses, and frame safeties would typically need a human to normalize everything for a VFX vendor—a luxury reserved for the biggest titles. With FDL, that mapping can be automated and controlled directly by users.

Inside the Media Production Suite

The Media Production Suite (MPS) lives inside Netflix Content Hub, built on a foundation of hybrid infrastructure. That infrastructure blends cloud-based and physically distributed capabilities across multiple global locations positioned close to productions. Local storage and compute connect through Netflix's Open Connect CDN to AWS, and Netflix is rolling out Content Hub Ingest Centers globally to ensure productions have sufficient upload speeds. Centralizing media removes physical transport, reduces human error, and improves both security and accessibility.

On top of that infrastructure, MPS provides a set of tools for managing media within the Netflix ecosystem:

  • Footage Ingest — uploads media and files into Content Hub.
  • Media Library — a central hub for searching, previewing, sharing, and downloading media.
  • Dailies — an operationally backed workflow that automates quality control, sound sync, color application, rendering, and delivery of dailies to editorial.
  • Remote Workstations — remote editorial workstations and storage for post-production.
  • VFX Pulls — automated conversion and delivery of VFX plates with associated color and framing files.
  • Conform Pulls — automated consolidation, trimming, and delivery of all OCF to picture-finishing vendors.
  • Media Downloader — an automated download tool that activates as soon as media is available in the Netflix cloud.

MPS aims to serve both advanced users and filmmakers with less experience. Individual tools are at different maturity levels, but over 350 titles have used at least one of them. Development input has spanned UCAN, EMEA, SEA, LATAM, and APAC, with contributions from Netflix teams, vendors, and global talent. The intent is to democratize technology: unlock wider talent pools, reduce non-creative work, and level the playing field across markets so more time goes into the creative work itself.

How Senna Put Netflix’s Media Production Suite Through Its Paces

Press enter or click to view image in full size

Media from the Brazilian-produced series ‘Senna’ being reviewed in MPS

The Brazilian series Senna—chronicling the life of Formula 1 driver Ayrton Senna—served as an early, high-stakes trial for Netflix’s Content Hub Media Production Suite (MPS). Shot across Argentina, Uruguay, Brazil, and the United Kingdom, with editorial teams in Porto Alegre and Spain and VFX work distributed among facilities in Brazil, Canada, the United States, and India, the production was a natural candidate for cloud-based workflow orchestration.

MPS operates on an opt-in model per tool, but the prerequisite for downstream services is uploading original camera files (OCF) and original sound files (OSF). For Post Supervisor Gabriel Queiroz, the geographic spread alone justified the approach: “to have all this material cloud-based, it’s definitely one of the most important things for us. It would be hard to bring all this media physically from Argentina or wherever to Brazil.” Uploading media on set eliminated the need to shuttle hard drives by air, create LTO tapes, or manage physical shipments—MPS does not require LTO tape writing unless a specific title demands it.

When Senna began production in June 2023, MPS was still in beta. Feedback from the production helped Netflix accelerate investment in the platform. Since then, ingest centers have been established worldwide where drives can be dropped off and OCF uploaded into the Netflix ecosystem within hours. The process is more involved than it appears: once a drive is connected and the Netflix Footage Ingest application is opened, a validation confirms all expected media from set is present. After upload, checksums verify media integrity, followed by inspection, metadata extraction, and creation of playable proxies for viewing, sharing, and downloading. Media is then automatically backed up to a second cloud storage tier for final archiving.

Previously, checking on these media management steps or confirming that camera cards could be cleared on set meant calling the post vendor. For Senna, anyone needing visibility could log into Content Hub and review activity in the Footage Ingest dashboard or look up details on past uploads.

Press enter or click to view image in full size

Remote monitoring media being uploaded and archived using the MPS Footage Ingest workflow

Though MPS offers services across the production lifecycle, Senna leaned heavily on VFX workflows. The show generated a high volume of footage and VFX shots, and with vendors spread across multiple countries, Queiroz called using MPS “basically a no-brainer.” He noted, “with the amount of vendors we have, knowing that we would have to deliver all this footage to all these kinds of vendors, including outside of Brazil and to different parts of the world,” a traditional manual approach would have been untenable. Manual VFX Pulls, the norm in Latin America, are prone to human error, slow for a production of this scale, and would have meant a distinct I/O method for every vendor.

With MPS, the Assistant Editor logged into Content Hub, uploaded an EDL, and the system automatically transcoded VFX Pulls, consolidated color files, and placed media into Workspaces—Google Drive-style folders built directly into Content Hub. The VFX Editor could adjust the directory before distributing shots to the appropriate vendors. Vendors returned shots to editorial or DI through the same MPS pipeline, giving the production a single standardized file transfer workflow rather than managing separate methods for each vendor.

Press enter or click to view image in full size

Illustrating a traditional VFX Editor having to manage various I/O methods

After picture lock, DI facility Quanta used MPS’s Conform Pull service for the online. Uploading an EDL triggered a QC of all referenced media to ensure a smooth conform, after which the system consolidated, trimmed, and packaged the necessary files. Since that early beta, improvements informed by Senna and other productions have expanded the system’s ability to match back to source media for both Conform and VFX Pulls. Instead of requiring an exact EDL-to-OCF match, several fuzzy matching variations are now possible, and Netflix is investigating perceptual matching algorithms to enable conforms based on computer vision rather than metadata alone.

Open Standards and the Road Ahead

MPS is built on open standards, which Netflix says makes the suite economically scalable and extends access to advanced production tools globally. The aim is to remove manual drudgery from the workflow so crews can concentrate on creative decisions. Cloud-based collaboration, the company argues, improves both efficiency and production quality. As the platform matures, the stated goal is to keep reducing friction for distributed teams and to broaden the range of productions that can operate with a globally connected, cloud-centric workflow.