The document pipeline behind legal platforms
Contracts, exhibits, and client recordings, virus-scanned, searchable, and processed in the region you pin.

Text found with OCR 
Attorney avatar 
Watermarked copy - PDFnda-signature-page.pdfJPGnda-signature-page.jpg

Converted to PDF nda-signature-page-page.webp · SHA-256
9e888e0e6e1f415adef3e7d804e022a382eccc6333d23cfd72a55e525df51586
Virus scanned
SHA-256 hash recorded - 1080p5.0 Mb/s720p2.8 Mb/s480p1.2 Mb/s360p0.6 Mb/s
Video ready to stream
Trusted since 2009 by
the world’s most ambitious platforms.
One pipeline for every case file
Prepare recorded interviews, check every client upload, and route productions, scans, and exhibits through reusable processing Templates.
Recorded interviews, ready to play
Encode recorded client calls and interviews for playback in your platform, with a poster frame and a draft transcript from the same upload.
Every client upload checked first
Virus-scan every file and refuse the types your intake flow does not accept, before anything reaches a reviewer.
Each file type gets the processing it needs
Productions, scans, exhibits, and recordings are routed by file type and source, each through the processing it needs.
- File
production-vol-01.zip
- Source
- Opposing counsel
- Routed through
- /file/decompress
- →/file/filter
- File
signed-nda.pdf
- Source
- Client portal
- Routed through
- /document/autorotate
- →/document/ocr
- File
exhibit-14.docx
- Source
- Case team
- Routed through
- /document/convert
- →/document/thumbs
- File
client-interview.mp4
- Source
- Case team
- Routed through
- /video/encode
- →/speech/transcribe
- File
bank-statements.pdf
- Source
- Client portal
- Routed through
- /file/hash
- →/document/ocr
- File
site-photo.jpg
- Source
- Investigator
- Routed through
- /file/hash
- →/image/resize
- File
client-records.zip
- Source
- Client IT
- Routed through
- /file/virusscan
- →/file/decompress
- File
contract-redline.docx
- Source
- Counterparty
- Routed through
- /document/convert
- →/s3/store
- File
site-visit.mov
- Source
- Investigator
- Routed through
- /video/encode
- →/video/thumbs
- File
court-filing.pdf
- Source
- E-filing
- Routed through
- /document/split
- →/document/thumbs
- File
foreign-letter.pdf
- Source
- Client portal
- Routed through
- /document/ocr
- →/text/translate
- File
closing-set.zip
- Source
- Case team
- Routed through
- /file/decompress
- →/document/merge
- File
financials.xlsx
- Source
- Client portal
- Routed through
- /document/convert
- File
scanned-will.pdf
- Source
- Client portal
- Routed through
- /document/ocr
- →/document/thumbs
- File
production-vol-01.zip
- Source
- Opposing counsel
- Routed through
- /file/decompress
- →/file/filter
- File
signed-nda.pdf
- Source
- Client portal
- Routed through
- /document/autorotate
- →/document/ocr
- File
exhibit-14.docx
- Source
- Case team
- Routed through
- /document/convert
- →/document/thumbs
- File
client-interview.mp4
- Source
- Case team
- Routed through
- /video/encode
- →/speech/transcribe
- File
bank-statements.pdf
- Source
- Client portal
- Routed through
- /file/hash
- →/document/ocr
- File
site-photo.jpg
- Source
- Investigator
- Routed through
- /file/hash
- →/image/resize
- File
client-records.zip
- Source
- Client IT
- Routed through
- /file/virusscan
- →/file/decompress
- File
contract-redline.docx
- Source
- Counterparty
- Routed through
- /document/convert
- →/s3/store
- File
site-visit.mov
- Source
- Investigator
- Routed through
- /video/encode
- →/video/thumbs
- File
court-filing.pdf
- Source
- E-filing
- Routed through
- /document/split
- →/document/thumbs
- File
foreign-letter.pdf
- Source
- Client portal
- Routed through
- /document/ocr
- →/text/translate
- File
closing-set.zip
- Source
- Case team
- Routed through
- /file/decompress
- →/document/merge
- File
financials.xlsx
- Source
- Client portal
- Routed through
- /document/convert
- File
scanned-will.pdf
- Source
- Client portal
- Routed through
- /document/ocr
- →/document/thumbs
What your reviewers and clients notice
The pipeline stays out of sight. What people on your platform see is a matter they can search, exhibits that open in place, and uploads that simply finish.
Scanned pages show up in search
Scanned PDFs are turned upright and read by OCR, so a search across a matter finds text that only ever existed on paper.
Raman1/1Exhibits open without a download
Word, Excel, and PowerPoint files become PDFs with page thumbnails, so reviewers read them in place.

exhibit-14.docxexhibit-14.pdfConverting to PDF…Converted from Word
Large productions upload once
Resumable uploads survive a dropped connection, and ZIP archives are unpacked and sorted by type the moment they land.
production-vol-01.zip4.2 GBUploading part 1 of 8Recorded calls arrive with a draft transcript
Client calls and interviews play in your platform alongside a draft transcript your team checks against the recording.
- Before we sign, can we go back to the term in section seven?
- Two years, with thirty days’ written notice to end it early.
- Good. Then we return all materials when the agreement ends.
I'm not usually one for fanboy-ism but Transloadit is one service I would recommend without hesitation. Their product and customer support is outstanding.What can Transloadit help with?
Six workflows legal tech and e-discovery platforms build on Transloadit. Each one is a Robot, or a recipe of them, already in our library.
Resumable production uploads
Resumable uploads for clients and opposing counsel sending gigabytes of documents, so a production survives a dropped connection.
Archives unpacked and sorted
Unpack uploaded ZIP and TAR archives, then filter the files inside by type, so each one gets the processing it needs.
Searchable text from scanned PDFs
Turn sideways scanned pages upright and extract their text, so scanned documents show up in your platform’s search.
Previews for every exhibit
Convert Word, Excel, and PowerPoint files to PDF and render page thumbnails, so reviewers read exhibits without downloading them.
SHA-256 checksums on record
A hash Step records a SHA-256 checksum of originals and outputs, so your platform can spot duplicates and check a file has not changed since.
Transcripts of recorded interviews
Draft transcripts of recorded client calls and interviews for internal review, ready for your team to check against the recording.
You keep the case.
We run the document pipeline.
Transloadit is the processing layer, not a replacement for your case management, review, or document management system. Here is exactly where the line sits, so nobody discovers it during integration.
What stays with your legal platform
- 01Matters, clients, permissions, and privilege decisions
- 02Review, coding, redaction, and Bates stamping
- 03Legal holds, retention, and deletion schedules
- 04Which documents are produced, and to whom
What Transloadit handles
- 01File upload or import and server-side processing execution
- 02Archive, document, OCR, hashing, and export steps
- 03Structured Assembly status and retryable notifications
- 04Exports of originals and outputs to the storage you configure
A document pipeline you do not have to maintain
OCR for scanned PDFs and images, and conversion for Office files, so every page in a matter reaches your search index.
Parallel processing turns a new production into searchable text and previews as it lands.
ISO 27001 and SOC 2 Type II, a DPA, signed uploads, and processing in the US, EU, or Asia Pacific.
Workers scale for discovery deadlines and bulk imports. Work beyond your plan’s job slots queues instead of failing.
Files export to your own S3, GCS, Azure, or R2. Temporary copies are purged after 24 hours by default.
Maintained SDKs for Node.js, Python, Go, and more, and workflows you version like code.
Legal document processing questions
The answers we would give on an evaluation call, including the cases where another tool is the better choice.
Can we send confidential client documents through Transloadit?
Start with nonsensitive sample documents. Before connecting client files, review your data requirements with our team, choose the processing region for each job, and export results to storage you control. OCR, transcription, and translation Steps call an external provider, so check that provider’s region against your clients’ requirements.
Does Transloadit redact documents?
No. Transloadit returns extracted text and previews; redaction, privilege review, and coding stay in your review tool. Once a redacted version is approved, it can run back through Transloadit for previews and exports like any other upload.
Which file types can clients upload?
PDFs, Word, Excel, and PowerPoint files, scans and photos, ZIP and TAR archives, and audio or video recordings. Filter Steps refuse the types your intake flow does not accept before anything is processed. Email containers such as PST and MSG are not parsed; extract them in your platform first.
Can we show that a file has not changed since it arrived?
Hash Steps record a checksum, such as SHA-256, for every original and output, and return it in the Assembly results. Your platform stores those checksums alongside its own audit trail.
What happens when a large production arrives before a deadline?
Workers autoscale with demand, so there is nothing to provision ahead of a peak. Your plan’s Priority Job Slots set how many jobs run at once; additional work queues instead of failing, and you can reserve extra slots ahead of a planned peak.
Does Transloadit keep copies of client documents?
Only briefly. Temporary processing copies are deleted after 24 hours by default, and you can turn off temporary result storage per workflow. Job status, which includes file names, is kept for 90 days by default. Originals and results are exported to the storage you configure, and retention and deletion happen in your own systems.
Is Transloadit an e-discovery platform or a document management system?
No. Transloadit is the processing layer between your clients’ uploads and your storage. Matters, review, and productions stay in your platform; Transloadit makes each file searchable, previewable, and scanned for known malware. See how document OCR works
Case files, on one API.
Run a sample set of nonsensitive documents through your first pipeline, and export searchable text and review previews to your own storage.