Candidate workloads
- Stable-schema JSONL and NDJSON
- Telemetry and generated event streams
- Workflow artifacts and repeated snapshots
- Structured tabular data with high repetition
Flagship 02 / ZJX
Store more history. Spend less to keep it.
ZJX is a Linux-first lossless archive system for telemetry, logs, workflow artifacts, repeated snapshots, and other structured data whose retention cost grows faster than its value.
RAW STRUCTURED HISTORY342.4MB
VERIFIED ZJX ARCHIVE4.41MB
98.7% smaller
The storage problem
Structured history is valuable. Repetition is expensive.
Operational data often repeats schemas, keys, values, and event shapes across millions of records. General-purpose archives can preserve it, but they do not always exploit that structure efficiently.
The result below is corpus-specific. It covers the complete retained archive and the strongest completed comparator in a 14-baseline run.
Measured archive size
Selected structured telemetry corpus
Raw input342.4 MB
Strongest completed comparator7.13 MB
ZJX archive4.41 MB
- Raw reduction
- 98.7%
- Smaller than comparator
- 38.1%
14 / 14 baselines completed; archive test passed.
- 01Qualify workload
- 02Create archive
- 03Inspect and test
- 04Plan restore
Boundaries
Results depend on schema, repetition, metadata requirements, and operating environment. ZJX does not claim a universal win across every log, table, document, source tree, or media workload.
It complements backup, encryption, access control, retention, and disaster recovery. It does not replace them.
Validate before deployment
Bring one expensive archive lane.
Greyforge can measure ZJX against serious comparators, verify the retained archive, and return an evidence packet for technical review.