Jev judgments for governed agent memory
As of October 1, 2026, Heartwood Memory's Jev judge for recall ranking is available to hosted Team and Professional organizations that opt in; it is off by default. Heartwood Memory can rank recalled memories with an opt-in judge that uses Jev, TypeSafe AI's System One model. Checking AI-written memories against their sources before they're saved is planned and not available yet.
It is off by default and offered only to hosted Team and Professional organizations that turn it on in writing. Community, self-hosted and air-gapped deployments never call Jev, and Heartwood's local ranker runs alongside every judged recall as the fallback.
What the Jev judge does
Rank recall results
The judge scores how well each policy-cleared memory answers the query, so the most relevant memories come first. Heartwood's local ranker runs alongside every judged recall, and recall modes the judge doesn't cover always use the local ranker.
Check AI-written memories before they're saved (planned)
Planned, not available yet. Before an AI-written memory is saved, the judge will check it against the source passages it cites. A memory is accepted only when Heartwood's built-in check and the judge both find it supported. A contradiction or an unsupported claim is rejected, and any other disagreement goes to review. The judge can make the check stricter, never looser.
What is Jev?
TypeSafe AI describes Jev as its first System One model. Instead of writing text, it answers structured questions (a yes or no, a choice among options, or a score) with a probability for each answer. See TypeSafe's announcement and documentation for details.
How it fits Heartwood's governed recall
This is how the judge works for recall ranking today.
Policy filters first. Records the calling agent isn't cleared to use are removed before anything is ranked. This part ships today.
Your organization's switch, then the egress check. The judge runs only if your organization has turned it on, and only after Heartwood's egress check allows TypeSafe AI for these records.
All or nothing. If any candidate is above your egress ceiling or labeled as personal data, or the request matches a secret pattern, nothing is sent and the whole recall uses the local ranker. Before anything is sent, Heartwood also removes email addresses, phone numbers and street addresses, and it sends nothing if the text still looks sensitive.
The local ranker stands by. It runs alongside the judge. If Jev is slow or unavailable, recall returns in the local order.
The call is recorded.
explain_recallshows which ranker ordered the results, the egress decision, the model and TypeSafe request ID, a hash of what was sent, and a probability for each memory the judge scored. The same record is added to Heartwood's hash-chained audit log.Trial first, if you want it. Your organization can start with a trial in which the judge's scores are recorded but the local order is used. Your records are sent to TypeSafe AI during the trial too.
What leaves your environment, and when
Only for a hosted organization that has opted in, and only when every check above passes.
| What | Sent to a judge? | What Heartwood keeps |
|---|---|---|
| The recall query and the candidate memories being ranked (clipped), when every candidate is cleared | Yes, to TypeSafe AI (US), including during a trial. TypeSafe AI is the only outside processor the judge calls. | The explain record and an audit row: ranker, egress decision, model, request ID, a hash of what was sent, one probability per scored memory |
| An AI-written memory and the source passages it cites | Not yet. This check is planned. When it's available, the memory and its cited sources will go to TypeSafe AI on the same terms. | — |
| Records the calling agent isn't cleared to use | Never. They're removed before ranking | — |
| A recall where any candidate is above your egress ceiling or labeled as personal data | Nothing is sent. The whole recall uses the local ranker | The reason |
| A request that matches a secret pattern | Nothing is sent | A refusal reason, never the matched value |
| Organizations that haven't opted in; Community, self-hosted and air-gapped deployments | Never | — |
| The judge's answers | — | Used for ranking and checking only. Never used to train or fine-tune any model |
TypeSafe AI is named here because it receives your data once you opt in. Retention on its side follows its own terms: TypeSafe AI legal. Heartwood doesn't describe this as a zero-data-retention arrangement.
What stays the same
- The self-hosted core makes no Jev calls, with or without this integration.
- Recall without the judge works as it does today: policy filters first, the local ranker orders the results, and explain_recall shows why.
- Nothing is required of you. If your organization never opts in, nothing changes.
What we'll publish, and when
We haven't published any accuracy, speed or cost figures for the Jev judge. We'll publish results only after measuring them on Heartwood's pre-registered recall test, with how each number was measured shown next to it. Results will be linked from this page and from our measurement receipts.
Turn on the Jev judge for your organization
For hosted Team and Professional organizations. Tell us which agent and which recall you'd put behind the judge, and who approves the opt-in for your organization. We target a response within one business day.
Questions
Does Heartwood Memory work with Jev?
As of October 1, 2026, Heartwood Memory's Jev judge for recall ranking is available to hosted Team and Professional organizations that opt in; it is off by default. It uses Jev, TypeSafe AI's System One model, to rank recalled memories. Checking AI-written memories against their sources is planned and not available yet.
What does the Jev judge do?
Today, one job: it reorders the memories recall returns so the most relevant come first. A second job, checking whether a memory an AI wrote is supported by its sources before it's saved, is planned.
Is it on by default? Who can use it?
No, it's off by default. It's offered only to hosted Team and Professional organizations, and only after an organization turns it on in writing.
What data is sent to TypeSafe, and who processes it?
Nothing, unless your organization opts in. After that, the recall query and the candidate memories being ranked go to TypeSafe AI in the US, including during a trial. TypeSafe AI is the named processor. Nothing is sent if any record is above your egress ceiling, labeled as personal data, or matches a secret pattern, and email addresses, phone numbers and street addresses are removed before anything is sent. Retention on TypeSafe's side follows its own terms, and Heartwood doesn't describe this as zero data retention.
Can I use it with Community or self-hosted Heartwood?
No. Community, self-hosted and air-gapped deployments never call Jev. Recall there always uses Heartwood's local ranker.
What happens if Jev is slow or unavailable?
Recall still returns, ranked by Heartwood's local ranker, which runs alongside every judged recall.
I can call Jev myself. Why use it through Heartwood?
You can. Jev is a public API, and other memory tools offer Jev ranking too. Heartwood's part is the governance around the call: only memories your policy clears can be judged, the feature is opt-in per organization, each judged recall is recorded with the model that ranked it, and the processor is named.
Do I get Jev API access through Heartwood?
No. Heartwood uses Jev inside its own service. The integration doesn't give you access to Jev or TypeSafe's API.
How much does it improve recall?
We haven't published figures. We'll publish them only after measuring on our pre-registered recall test, with how each number was measured shown next to it.
Does Heartwood train models on Jev's answers?
No. Heartwood doesn't use Jev's outputs to train or fine-tune any model.
Is Heartwood Memory affiliated with TypeSafe AI? Is it connected to other Jev memory projects?
No to both. Heartwood Memory is made by Edukas Solutions LLC and is not affiliated with or endorsed by TypeSafe AI. Third-party Jev memory projects are not made or endorsed by Heartwood Memory; its own Jev integration is the one described on this page.
How do I turn it on, and what does it cost?
Use the form on this page. We turn it on for your organization after a written opt-in, and you can start with a trial. It's included in Team and Professional.
Jev is a model from TypeSafe AI, Inc. Heartwood Memory is made by Edukas Solutions LLC and is not affiliated with or endorsed by TypeSafe AI.