Amazon S3 to Premiere Pro, with Imaginario’s video intelligence in the middle
Search years of footage where it already lives in S3, by what is said and what is on screen. Cut it, then send clips, sequences and timecoded metadata straight into Premiere Pro or DaVinci Resolve. No migration, no second archive.
The workflow Warner Bros. Discovery EMEA runs for Cartoon Network and Boomerang.
The workflow
In large media companies the content ultimately lives in cloud storage, and Amazon S3 is the most common home for it. Imaginario connects to the buckets you choose, indexes them from proxies (3 to 5 Mbps, any resolution, encoded from the originals if none exist), and gives marketing, corporate communications and post-production teams one place to search years of footage by what is said and what is on screen.
The moments they find go into Premiere Pro as clips, sequences and timecoded metadata, or into DaVinci Resolve and Avid Media Composer. Nothing is migrated and no second archive is built.
This is the cloud video editing workflow Warner Bros. Discovery EMEA has run for four years across Cartoon Network and Boomerang: content in S3, search and clipping in Imaginario, finishing in Premiere Pro, with almost 70% less time spent searching and clipping for social media. Read the Warner Bros. Discovery story.
What Imaginario does in the middle
Imaginario is the intelligence layer between the storage and the edit. It never replaces either. It understands the footage at shot level, what is said, who is on screen, what happens and how the story unfolds, so it can surface highlights, answer questions about the library, assemble rough cuts and return insights and metadata. That is more than the tagging you get from a MAM add-on, and it is what the three steps below are built on.
| Find | Video search across everything indexed: dialogue search for what is said, visual search for what is on screen (including image-to-video search), audio search for sound, and person search for who appears. Ask your library answers questions across the whole archive in plain language, and AI indexing produces the timecoded metadata behind all of it, in more than 100 languages and dialects. |
| Cut | StoryLab turns the moments you find into single clips for social and multi-video stories for compilations and rough cuts, with Highlight to Edit for a first pass and Clip Studio to trim, reframe, brand and subtitle. Captions, transcripts and chapters come from the same index. |
| Deliver | Export a clip or a whole collection as a sequence with timecoded metadata into Adobe Premiere Pro, DaVinci Resolve or Avid Media Composer, send finished clips back to S3, return metadata to a MAM or DAM, or publish straight to YouTube and Shorts. |
What goes in and what comes out
| Direction | What moves |
|---|---|
| From Amazon S3 | The buckets and prefixes you choose, indexed selectively from proxies at 3 to 5 Mbps in any resolution, existing or encoded by Imaginario from the originals |
| Into Premiere Pro (or Resolve, Avid) | Clips, sequences and timecoded metadata, ready to edit |
| Back to S3, your MAM or your channels | Finished clips to S3, time-based and asset-level metadata to a MAM or DAM, published cuts to YouTube and social platforms |
| What stays put | The originals, your storage classes, your IAM policies |
How to set it up
| Step | What happens |
|---|---|
| 1. Connect the buckets | Self-serve: in Imaginario, add the bucket name and the access key and secret key your cloud team creates for it, then choose the prefixes to index. Add as many buckets as you need. Start with one franchise or one campaign and add more later. |
| 2. Let Imaginario index it | Imaginario works from proxies at 3 to 5 Mbps, at any resolution. If your buckets already hold proxies it uses those, and if they hold only originals it encodes the proxies itself. It then indexes what is said, what is on screen and who is on screen, including animated characters. Speech is analyzed in more than 100 languages and dialects, and the search engine works natively in more than 17 languages, so a team in Paris or Istanbul searches in its own language. |
| 3. Search in plain language | Describe the moment: a line, a character, a location, an action. Filter the results, build collections, and finesse the moments in Clip Studio with branding and subtitles. |
| 4. Build the cut | Use StoryLab to turn the moments into single clips for social or a multi-video story for a compilation, or hand-pick them into a collection for the editor. |
| 5. Send it to Premiere Pro | Export a clip from one asset, or a collection across many assets as a sequence for compilations and rough cuts. Timecoded metadata travels with every clip. The same export goes to DaVinci Resolve or Avid Media Composer. |
| 6. Publish or write back | Finished cuts go back to S3, into your MAM or DAM as time-based and asset-level metadata if you run one, or straight to YouTube and other social platforms. |
Three ways teams use it
| Use case | How it works |
|---|---|
| Franchise libraries for social | A brief arrives phrased by character, not by episode: every time a character says “fun” or laughs, every Halloween moment across the slate. The team searches the whole franchise library in S3 with dialogue, visual and person search, cuts the moments as single clips in StoryLab and Clip Studio, and sends the sequence to Premiere Pro. That is how Cartoon Network UK grew from 10 million YouTube subscribers in early 2024 to more than 11.8 million. |
| Seasonal compilations from the whole slate | Compilations, anniversary edits and sponsor packages draw on older titles and secondary show brands that nobody could find before. Multi-video stories assemble them across every bucket, and the long tail of the catalogue comes back into rotation. |
| Corporate communications and post-production rough cuts | Corporate communications teams find the CEO quote during a corporate event, the customer testimonial, the product shot or the event moment across years of archive in S3 and publish it the same day. Editors search the same library, build a selection across many assets and export it as a sequence. One index, several teams, no duplicated storage. |
Who runs this workflow
Large media companies and enterprises where marketing, corporate communications and post-production share one cloud archive, and agencies managing a client library in S3.
Integrations in this workflow
Source: Amazon S3, and with the same steps Google Cloud Storage, Microsoft Azure, Backblaze, Wasabi. Destination: Adobe Premiere Pro, or DaVinci Resolve and Avid Media Composer. See all cloud storage integrations, the Warner Bros. Discovery story and the other popular workflows.
Frequently asked questions
Do I have to copy my S3 buckets somewhere else?
No. Imaginario connects to the buckets and prefixes you choose, pulls lightweight proxies and indexes them. The originals stay in Amazon S3 under your account, your storage classes and your access policies.
Which teams use this workflow?
Marketing, corporate communications and post-production teams, and in large enterprises all of them on the same library. Warner Bros. Discovery EMEA runs it for Cartoon Network and Boomerang: video marketing and branding in London and the teams in France, Italy, Spain, Turkey, the Middle East and Africa, plus post-production. Read the Warner Bros. Discovery customer story.
Does the workflow need a MAM?
No. Creative teams rarely work inside a MAM, which serves content operations and engineering. When the content lives in S3, Imaginario plugs into the storage directly. If you do run a MAM or DAM, Imaginario can return time-based and asset-level metadata to it as well, mapped to your fields with taxonomy and schema mapping.
What arrives in Premiere Pro, and does it work with other cloud storage?
Clips, sequences and timecoded metadata, ready to edit, from a single asset or across many assets as a sequence (walkthroughs for single-asset and multi-asset export), and the same export goes to DaVinci Resolve and Avid Media Composer. The same steps apply to Google Cloud Storage, Azure Blob Storage, Backblaze and Wasabi.
Get started today
Enjoy thousands of minutes of AI analysis, transcriptions, search, and auto-captions, plus much more.