Jylus
give AI systems evidence from changing data.
Last verified:
What is Jylus?
Jylus is a data and evidence engine for AI applications that converts changing records into compact, source-backed Context Packs containing facts, timeline evidence, contradictions, missing evidence and proof IDs. It targets developers building AI agents and real-time data systems who need to give models verified, bounded context instead of raw documents.
Jylus pricing
Pricing model: Freemium
Free tier with 1 GB storage, free API key (no credit card); paid plans not shown on this page
Jylus pros
- Returns source-backed evidence with proof IDs, conflicts and missing-evidence flags rather than final answers
- Reported token reduction (e.g., 76% in example) via bounded context packs with token budgets
- Supports temporal queries separating 'what happened' from 'what was known at a time'
- Free single-use trial with no account required and isolated, auto-deleted tenants
Jylus cons
- Does not generate final answers — only compiled evidence, requiring integration work
- Trial ingestion limited to 32 KB text/JSON/log files
- Data region limited to Sydney, Australia; free tier capped at 1 GB storage and 1,000 events/sec
Frequently asked questions about Jylus
Is a credit card or account required to try Jylus?
No. A single-use trial runs without an account in an isolated one-time tenant that is deleted afterward, and a free API key requires no credit card or sales call.
Does Jylus answer questions directly?
No. Jylus returns evidence and source proofs for your model; it does not generate the final answer.
Is submitted data used for training?
The page states your data stays yours and is never used to train general-purpose AI.