Data governance middleware
Filcerne sits between your source systems and everything downstream. Fields an administrator hasn't approved don't move forward — they're held back, logged, and never reach the other side.
Built for banking · retail · healthcare · insurance
How it works
Every record takes the same path, whatever format it arrived in and wherever it's headed.
Records arrive from any of 17 sources — a Kafka topic, an S3 drop, a database query, a REST call.
When one entity ID arrives under two near-matching spellings of a name, both records still move. The discrepancy goes to a report you can act on.
Your business rules run. Missing a required field, or resubmitting an unchanged profile, produces a coded rejection.
Every field is checked against the allowlist. Approved fields pass, PII and PCI columns encrypt, everything else is held.
Cleared fields go downstream in whatever format that system needs — CSV, Excel, XML, Parquet, JSON.
Where it sits
Both matter. They are not the same control, and they do not sit in the same place.
Access governance platforms
Governs data in transit
If your exposure is an analyst over-reaching in a warehouse, buy an access governance platform. If it is a nightly feed quietly carrying nine fields a downstream partner was never approved to receive, that feed never touches a query engine — and that is the gap Filcerne is built for.
Sources and destinations
Input format and output format are set separately, so neither side has to change to work with the other.
What it does
Most tools decide whether a whole record moves. Filcerne decides column by column.
An administrator lists what may leave. Anything not on that list is held back by default — including fields nobody knew were in the feed.
Flag a column and it travels as ciphertext. If the key is ever missing, the field is held rather than sent in the clear.
Near-matching names under one ID are flagged for reconciliation. Filcerne never picks a winner or rewrites a value — the source systems stay the record of truth.
Every blocked record gets a code and a plain-English reason, searchable by entity, name, or date for 30 days.
Built for regulated data
Every choice assumes the data is sensitive and the configuration might be wrong.
Fields are held unless explicitly approved, so a new column upstream can't quietly start flowing downstream.
A field marked for encryption with no key available is withheld — a misconfiguration can't turn into a leak.
The audit trail records field names and salted hashes, never the values, so the log itself isn't a second copy of your PII.
Plain-HTTP destinations are refused outright. Downstream endpoints have to be TLS.
Reading data is one permission. Reading logs, exporting reports, and decrypting values are another.
Filcerne is early. We are pre-SOC 2 and run pilots on masked or synthetic data while certification and independent penetration testing complete. We would rather tell you that now than have your security review find it.
Export what was delivered, rejected, held, or flagged for reconciliation as CSV or Excel — for an auditor, a regulator, or a Monday morning question.
About 30 minutes. Send us your field names — no data — and we’ll show you exactly what would and wouldn’t reach your downstream systems.