Territory scoring now completes more reliably for teams turning account research into a ranked plan. A previously failing step in the scoring workflow has been corrected, so eligible scoring runs can finish and save their results instead of stopping during processing.

What changed

Previously, a scoring run could fail after it had started, which could leave a territory plan waiting for results or require the team to try again. Enki IQ now completes the affected update steps safely as scoring work progresses.

The fix also covers related territory-plan progress updates that use the same workflow, keeping scoring activity and plan status aligned as the work moves forward.

Why it matters

Scoring is how teams turn a large account set into a practical starting point for territory planning. A more dependable run means less interruption between selecting accounts, reviewing scores, and moving into the next planning step.

Details

• Applies to territory-scoring workflows for organizations using Enki IQ.

• Scoring results and related plan progress now continue through the corrected processing path.

• No setup or workflow change is required.

People who have already been verified for access now stay in Enki IQ if a later routine access check briefly cannot be completed. This prevents a temporary network interruption from unnecessarily replacing an active session with an access-verification screen.

What changed

Previously, a later access refresh could interrupt an already-authorized session if the browser could not complete the request. Enki IQ now keeps the verified session available in that situation and retries eligible read-only checks once with refreshed sign-in information.

Initial access verification remains protective. If Enki IQ cannot verify access when a session first starts, it continues to require a successful check before showing protected workspace content.

Why it matters

Teams can keep working through short-lived connectivity issues instead of losing their place in the application. At the same time, first-time access decisions remain subject to the same safeguards that protect organization data.

Details

• Applies after Enki IQ has successfully verified access in the current session.

• Later transient check failures no longer block an entitled workspace.

• Initial verification failures still prompt the user to retry or sign out.

Organization onboarding can now recover more reliably when an automated company-profile draft does not finish as expected. Teams setting up an organization are less likely to be left waiting for a draft after an interrupted background step, while the system continues to protect against duplicate work.

What changed

Previously, a draft that became safely stale could require someone to return and restart the onboarding action. Enki IQ now checks for eligible stalled drafts and can recover them without requiring the person who started setup to stay on the page.

The recovery flow uses the same safeguards as the initial request: it avoids replacing work that may still be active and does not run multiple drafts for the same setup action.

Why it matters

Onboarding is the first step to using Enki IQ with your team. This update makes that step more dependable when a browser session ends, a connection is interrupted, or a background task needs to be safely retried.

Details

• Applies to organization setup flows that create a company-profile draft.

• Recovery is automatic when a draft is confirmed safe to replace.

• No action is required from administrators or new users.

Revoked API keys are now denied access on the next request. This gives administrators a dependable way to remove a credential from use when it is no longer trusted.

What changed

Previously, a key could remain usable briefly after it was revoked. Enki IQ now confirms a verified key’s current status before authorizing API-key access. Revoked, expired, missing, or otherwise unavailable credentials are refused instead of being allowed to continue.

Why it matters

Revoking a key is a time-sensitive security action. This update closes the gap between that decision and its effect, helping teams contain a lost, exposed, or no-longer-needed credential with greater confidence while keeping normal access rules intact.

Details

• Immediate effect: A revoked key is refused on the next API-key request.

• Existing safeguards remain: Expired and invalid credentials continue to be refused, while valid supported keys keep working.

• No action required: The protection is active automatically for API-key authentication.

We upgraded the release safeguards behind Enki’s AI assistant after a deployment capacity transition took longer than expected to settle.

• Automatic recovery: If warm capacity does not stabilize, the release process now restores the previous live version, traffic allocation, and capacity settings as one coordinated rollback.

• Safer completion checks: Releases wait for capacity to reach a healthy state before they are marked complete.

• Independent verification: Post-deployment checks confirm the assistant is serving from a single healthy version and that routing and authentication settings remain unchanged.

• More accurate monitoring: Low-traffic windows with no metric samples are handled explicitly, preventing an empty monitoring interval from being mistaken for an outage.

The result is a more predictable release path for the AI assistant, with faster recovery and clearer evidence that service health was preserved.