Your data is yours. In any shape your stack speaks.
Every order, product, customer, every file your team uploaded. Export it on a schedule. Query it in your data warehouse. Send it to retailers in their format. Pipe it to your ML team. Walk out the door with all of it whenever you want.
The lock-in story inverted.
Every other platform: your data lives in their tables, in their format. Cancel the bill, lose the data. That asymmetry is the business model.
CCEN starts from the inverse premise. Your data is yours. Every CCEN entity (orders, products, customers and the rest) exports to Parquet, the format every modern data warehouse speaks. Every operational table is open to your analysts. Every file your team uploaded is downloadable by URL. EDI and SFTP for retailers, REST and GraphQL for engineering.
What that buys you: train your own ML on your real customer history. Pipe a fresh export to your warehouse every hour. Hand your analysts a SQL prompt and let them go. Plug in a BI tool that bypasses CCEN entirely. CCEN earns its keep by being useful, not by holding your data hostage.
Seven entities. Universal by design.
Every commerce business has these seven things. CCEN models them the same way for everyone. That shared shape is what lets every app on top, ours or third-party, work with the same data.
Line items, fulfillments, channel-unified, bundle-aware
Variants, inventory levels per location, internal source of truth
Per-channel projections of products, copy/compute/custom fields
Records, segments, LTV, history, contact data
Calendar of holidays, sale events, launches, campaigns
Builder engine, saved reports, dashboard tiles
Every mutation, every actor, every timestamp
An order item links to a listing variant, not directly to a product variant. This is the one design choice that makes everything else work. Bundles work because a single listing maps to multiple products. No painful workarounds, no per-channel hacks, no spreadsheet at 2am.
Four shapes. No connector tax.
Parquet
The format every modern data warehouse speaks (BigQuery, Snowflake, Redshift, Databricks). Schedule snapshots or live streams of every entity to your S3 bucket. No paid connector tax.
DuckDB
Run the same SQL we run, locally, from a notebook or a script. Zero infrastructure. The query plan is the same plan the platform runs internally. Ideal for ad hoc analysis and ML feature engineering.
EDI and SFTP
The retailer feeds Walmart, Target, Macy's, Costco, TJX, Kohl's and Nordstrom expect. Standard EDI documents (850 purchase orders, 855 confirmations, 856 ship notices, 810 invoices, 940/943/944/945 warehouse) pre-mapped. Daily SFTP drops with retailer-specific quirks handled.
REST and GraphQL
Modern APIs your engineering team will recognize. Same API our own apps use. Webhooks push to you when things change, so you don't have to poll.
-- 90-day GMV by customer, no ETL pipeline
SELECT
customer_id,
COUNT(*) AS orders_90d,
SUM(net_total) AS gmv_90d
FROM read_parquet('s3://your-bucket/ccen/orders/*.parquet')
WHERE created_at >= NOW() - INTERVAL 90 DAY
GROUP BY 1
ORDER BY gmv_90d DESC
LIMIT 100;# ccen.exports.toml
[orders]
format = "parquet"
schedule = "*/15 * * * *" # every 15 minutes
target = "s3://your-bucket/ccen/orders"
partition = ["created_date", "channel"]
[products]
format = "parquet"
schedule = "0 * * * *" # hourly
target = "s3://your-bucket/ccen/products"Every file accessible.
POs, packing slips, ASN PDFs, design files, supplier invoices, uploaded photos for listings. CCEN's file layer is browsable, searchable, and exportable. Files live behind signed URLs with audit-tracked access. The directory tree is yours.
No more "we lost that PDF when we switched WMS providers." Every attachment, every revision, every actor on it.
/orders
/1047-8832
receipt-2025-03-14.pdf 62 KB
packing-slip-LA.pdf 44 KB
proof-of-delivery.png 218 KB
/purchase-orders
/PO-2451
PO-2451-supplier.pdf 128 KB
invoice-supplier.pdf 96 KB
ASN-856.edi 12 KB
/products
/MBG-DRS-034-PNK
hero-front.tiff 3.4 MB
swatch-detail.jpg 412 KB
/audit
access-log-2025-03.csv 9.8 MBCompliance that does not get in the way.
Residency
Pin your tenant to US, EU, or APAC. Backups stay in-region. We can split tenants by region for global brands.
Retention
Configurable per entity. Default infinite for orders and audit. Per-entity retention schedules for PII-heavy tables.
Audit logging
Every change, every person, every timestamp, every IP. Export it. Stream it to your security team's tools (SIEM-ready, if your security team uses that word).
Encryption
Encrypted at rest and in transit (AES-256, TLS 1.3). Each customer gets their own encryption key, rotated on schedule. SOC 2 Type II audited.
The exit guarantee.
If you ever decide to leave, the data leaves with you. No migration tax, no "professional services engagement," no hostage negotiation. The export pipelines are part of the product, not a paid add-on.
Talk to an engineer- Full Parquet snapshot
Every L0 entity, every L1 app's table, exported to your bucket within 24 hours of a request.
- File-system tarball
Every uploaded file, every PDF, every image, packaged with manifest and SHA-256 checksums.
- Audit log preserved
Full audit log included. Every change, every person who made it, every timestamp. Exported as Parquet.
- API kept live
Read access stays live for 90 days after you cancel, so your next system can ingest cleanly. We don't make leaving painful on purpose.
- No re-ingestion fee
We do not charge for the export, the SFTP transfer, the storage costs, or the engineer time.
Stop stitching tools together. Start running your business.
See CCEN with your own data in a 30-minute call. No slides. Just the product.