DATA MANAGEMENT
Files you moved once,
as data you can use again.
Link files scattered across servers, object storage and on-site devices into a single Dataset,
and follow one file's route and change history exactly as it happened.
DATASET
Scattered files,
as one Dataset.
Files sitting on different servers, object storage and on-site devices are linked into a single Dataset and managed together, without being moved.
A Dataset is a logical set of pointers to where the files already are. It does not copy them into one place.
ServerSeoul HQ
- design.zip
- spec_document.pdf
- project_plan.xlsx
Object StorageS3 Storage
- model_data.bin
- media_assets.zip
- training_set.tar
Edge DeviceFactory Line
- sensor_logs.csv
- vision_data.mp4
- inspection.zip
Remote SiteTokyo Branch
- reference.pdf
- manual.zip
- daily_report.xlsx
Production DatasetFiles across multiple locations · One logical dataset
DISTRIBUTED DATASETS
Wherever the files are,
manage them as one.
Search files held on servers, object storage and on-site devices from a single screen, and manage all of them the same way.
Every location on one screen
Files on different servers, storage services and on-site devices are reviewed and managed inside one Dataset.
The same way, wherever they live
Local servers, object storage, edge devices — the storage differs, the interface for searching, selecting and managing does not.
Ready for what you add next
A new device or storage service joins an existing Dataset and carries on being managed with the rest.
Datasets
- Production Assets1,248 files · 3 locations
- Training Data8,420 files · 2 locations
- Media Archive2,193 files · 4 locations
- Daily Collection365 files · 5 locations
- Design Files542 files · 2 locations
Locations in this dataset
- Seoul HQ / Server532 files · 42%
- S3 Storage428 files · 34%
- Factory Edge288 files · 23%
DATASET REUSE
A file moved once,
used again in the next job.
Files that arrived by collection, transfer or web upload are kept as a Dataset
and handed straight to the next transfer job as its input.
Collect or transfer
Factory CollectionFactory Line → CenterCompleted
- machine_log_001.csv2.1 GB
- machine_log_002.csv2.3 GB
- inspection_image.zip8.4 GB
- sensor_data.tar12.6 GB
1,248 files collected
Keep as a Dataset
Production Dataset1,248 files · 3 locations
Reuse in the next job
New FlowCenter → GPU Cluster
Source
Production Dataset
Destination
GPU Cluster
- Changed files only
- Verification after transfer
- Schedule
Run Flow
Still there for the job after that
Collected results as they are
Files collected from equipment become a Dataset straight away.
Web uploads too
Files uploaded through EX are managed exactly the same way.
Reused across jobs
One Dataset feeds as many transfer jobs as you need.
Always current
Pick only the files that changed and send those to the next job.
FILE TRACE & LINEAGE
Down to one file's route,
all of it on record.
Where a file was created, when it moved and which job carried it — the whole route is traceable from one screen.
Search results (5)
- model_weights.bin/datasets/ai-training/48.6 GB
- model_weights_v2.bin/datasets/ai-training/48.6 GB
- model_config.json/datasets/ai-training/12 KB
- AI Training Dataset3 files · 78.6 GB
- Daily Collection365 files · 5 locations
model_weights.bin48.6 GB · AI Training DatasetLast updated 2025-10-19 14:22
File Journey
- Source
- Seoul HQ / Server
- Destination
- Data Center
- Transfer result
- Completed (Verified)
- Transfer time
- 12m 34s
- Average speed
- 5.9 GB/min
- Verification
- SHA-256 · Matched
FILE HISTORY & METADATA
The file may change.
The record never breaks.
Metadata, change history, transfer results and verification status are all reviewed and managed from one screen.
File information and metadata
Basic file information, the Dataset and project it belongs to, business tags and more.
Change history
Everything that happened to the file — created, moved, modified — in order.
Transfer result and verification
The result of each transfer, how long it took and whether it verified: a record you can rely on.
inspection_data.zip8.4 GB · Factory Collection Dataset
Latest Transfer ReceiptVerified
- Transfer type
- Server to Server
- Source
- Factory Edge
- Destination
- Data Center
- Transferred size
- 8.4 GB
- Duration
- 12m 14s
- Checksum (SHA-256)
- 3f8a2c6e9d…7b1e4d9a3c
- Verification
- Passed