Xurrent ITSM - July 2026 Product Updates

This post is a living document, updated throughout the month as new product releases roll out. Each update listed here is released to our QA environment on the date shown and typically promoted to production the following week. Along the way, customers can review changes, share feedback, and help shape what ships next. Xurrent delivers product updates on a weekly cadence to keep enhancements moving continuously.

July 9, 2026

CI Reconciliation: Readable Proposed Changes for Custom and Rich-Text Fields

When a configuration item change lands in the reconciliation queue, the "Proposed Changes" panel now presents the difference the way the rest of the product displays records.

Each custom field that actually changed appears on its own row with its configured label and a human-readable value. Remarks and other rich-text fields render as formatted text with working inline-image previews, including images added as part of the proposed change. Fields that did not change no longer appear as emptied. When an approval cannot go through unchanged, a reviewer can edit the proposed remarks directly in the rich-text editor before approving.

For privacy, the raw bundle of custom-field data is no longer sent to the browser. Approval always applies the value stored with the proposed change.

Sera AI: Configurable Chat Widget Name

Account administrators can now set the name shown in the Sera AI chat widget header on the Self Service homepage. A "chat box display name" field appears in the "Global Settings" tab of "Self Service Design", in the "Self service homepage" section directly below the "Show chat widget" control.

Setting a value replaces "Sera AI" in the widget header for end users; leaving it blank keeps the header reading "Sera AI". The field accepts plain text up to 40 characters, trims whitespace, and treats a whitespace-only value as empty.

The override affects the widget header title only; supporting strings in the widget are unchanged. The value is display-only and does not change the agent's internal name, instructions, persona, or how interactions are logged. It applies only to the account's own self-service experience. With "Show chat widget" unchecked the field has no rendered effect, and an entered value is preserved if the widget is unchecked and rechecked.

Self Service: 'Submit on Behalf Of' Available Regardless of Portal Layout

The new request wizard now offers the "submit on behalf of" option on every step of the flow, in any account that permits submitting on behalf of someone else. Availability depends only on whether the account allows it, not on the portal's layout, so users whose portal opens directly on a service instance or a request template can now reach it.

This builds on the earlier fix on the same request that clears the self-registered flag when an account administrator reassigns a person to a different organization.

Accessibility: Ongoing Conformance Updates

This release includes several accessibility conformance improvements identified through recurring monitoring. Anchor elements that carry an "onclick" handler but no "href" now automatically receive role="button" and tabindex="0" so they are reachable and operable by keyboard and assistive technology. The page's html tag now sets its lang attribute dynamically to match the active language. Positive tabindex values have been removed so focus order follows the natural document order, and legacy ".button" class references have been cleaned up.

We review accessibility conformance on a recurring basis, and further updates will follow in future releases.

July 2, 2026

Sera AI: Closure Agent and Triage Agent

Xurrent Agents are now available in Sera AI. Two agents ship in this release: Closure Agent and Triage Agent. Account administrators manage agents in a new Agent Designer section of the settings console, and each agent has a set of configurable skills.

Closure Agent manages requests around the resolved to closed lifecycle. Its skills

• Verification Note Generation drafts a customer-facing note when a request reaches a resolved or completed state, summarizing what was done and inviting the requester to confirm.
• Reopen Signal Detection judges whether a requester's reply after a request was resolved signals the issue was not actually fixed. In "Suggest" mode it posts an internal note asking the developer or specialist to reopen and complete the fix; in "Auto Confirm" mode it reopens the request and reassigns it back to that developer.

Triage Agent acts on requests after the auto-classifier sets the team. It applies team-specific intelligence and does not handle team assignment. Its skills:

• Duplicate Detection identifies whether an incoming request is a duplicate of an existing open request on the same team.
• CI Linking suggests Configuration Items relevant to a request based on its subject, description, and historical patterns.
• Impact Assessment estimates the business impact of a request based on affected CIs, user role, and historical incident patterns.

Skills operate in either "Suggest" mode (posts a recommendation for a human to review) or "Auto Confirm" mode (performs the action directly). Each skill can be tuned per team based on the level of automation the team wants.

Adjustments can also be made to the confidence threshold at which the agent can take a given action.

An agent can be added to a team like any other member and then assigned to individual requests by a human team member.

In this release, agents are assigned to requests manually by team members. Automations for assigning agents based on rules or triggers are planned for a follow-on release.

Sera AI Studio: Knowledge Article Readiness

Sera AI Studio now includes a "Knowledge Article Readiness" tab, sitting alongside Agent Instructions, Golden Set, Template Readiness, and Settings. It gives you a single place to see how ready your knowledge articles are for Sera to use, and what is holding any of them back. The layout matches Template Readiness, so if you have used that tab, this one will feel familiar.

Four summary cards across the top give you the headline picture: "Total articles", "Enabled" (articles Sera is allowed to draw from), "With gaps", and a "Readiness score". The readiness score shows the percentage of enabled articles that are fully configured, with an N of M caption and a progress bar so you can track progress at a glance.

Each article is checked against four fields: "Service", "Keywords", "Description", and "Instructions". A field shows "Present" when it is filled in and "Missing" when it is empty. "Service" matters most here. Sera retrieves knowledge within the services it covers, so an article with no service attached will not surface for customers no matter how good the content is.

A "Completeness gaps" table lists only the articles that have at least one missing field, sorted so the articles needing the most attention appear first. Click an article title to open it, fill in what is missing, and watch the readiness score update as you go.

The score is based only on the articles Sera can actually use: published, non-archived articles within the services Sera covers. That means it reflects the real state of the knowledge Sera is working with, rather than being diluted by content the agent never sees. Because the score is calculated the same way as Template Readiness, you can compare the two directly.

Analytics: Search Phrases Now Include Sera AI Chat

The Search Phrases report at Analytics > Search Phrases now includes a "Sera AI Chat" tab, alongside the existing "Self Service" and "Service Desk" tabs. Account Administrators can see what topics drive end users into Sera AI conversations.

Each Sera AI conversation is summarized into a concise search phrase per intent, rather than logged as raw messages. A conversation that spans more than one distinct ask can produce more than one phrase. Personal data (email addresses, phone numbers, names, account or ID numbers) is detected and obfuscated before the phrase is stored or shown.

Sera AI Chat phrases are ranked by frequency, consistent with the other tabs. The Conversions and Conversion percentage columns are hidden on this tab, since chat has no result list or conversion rate; both columns remain in place on Self Service and Service Desk.

The tab is shown only for accounts with the virtual agent enabled.

Shop: Bundles Now Available

You can now group shop articles that belong together into a single bundle. For example, a New Hire Kit that includes a laptop, monitor, and headset. A shop administrator sets the bundle up once, and requesters order the whole thing as one line in the shop.

Setting up a bundle

Bundles live in their own "Shop Article Bundles" record. You give the bundle a "Name" and fill in the usual shop article fields: "Product", "Category", "Short description", and "Full description". Leave "Disabled" unchecked to make it available.

Then, under "Bundle Items", click "Link items..." to add the articles the bundle contains and set a quantity for each. Below that, "Pricing" and "Availability" have their own sections, covered next.

One thing to keep in mind while adding items: every article in a bundle must share the same pricing cadence, either all one-time or all recurring at the same frequency. The first item you add sets the cadence, and the picker then shows only matching items. (Bundles cannot contain other bundles in this release.)

Pricing

You can let a bundle price itself, or set a fixed price. If you leave the price field empty, the bundle price is the total of its contained articles, calculated in each account's own currency. If you enter a price, that price is shown as-is, with no ceiling relative to the item total. If you set the price to zero, the bundle shows no price at all.

Keeping bundles current

A bundle turns off automatically if any article inside it is disabled or removed, and an audit entry records why. It turns back on automatically once all its articles are valid again. You only need to step in when an article is permanently removed and leaves a gap the bundle can no longer resolve.

If you have set a fixed price and the price of an item inside the bundle changes, the bundle stays active, but a fixed price does not update on its own. You will be warned so you can adjust it yourself. And if you edit a standalone article in a way that would take its bundles offline, you will see a confirmation listing which bundles are affected before the change is applied.

What requesters see

Bundles available to a requester's account appear in the shop catalog and look like any other article. Coverage still applies per item, based on the requester's Service Offering: a self-priced bundle drops any items the requester is not covered for and prices from what remains. A fixed-price bundle only appears if the requester is covered for every item in it. A bundle the requester is covered for none of is hidden either way.

Cart and ordering

In the cart, a bundle is a single line. When the requester submits, it becomes one request per covered item, each routed under its own Service Offering and fulfillment template, exactly as if the item had been ordered on its own. Every request that comes out of a bundle records the bundle name for reference. The bundle itself is not a request.

API, import, and export

The shop article REST and GraphQL APIs, the audit trail, and CSV import and export all cover the bundle flag, the contained items and their quantities, and the custom bundle price.

Request Templates: Subject Translation Opt-Out

Request templates now support an optional parameter so that the subjects of requests created from it are never auto-translated. This is useful for templates whose subjects are structured operational identifiers (rollout codes, branch numbers, asset numbers) where translation can corrupt the intended value.

A new "Translate subject" checkbox appears in the subject group on the request template, beneath the Subject field, alongside "Disabled" and "Copy subject to requests". It defaults to checked, preserving today's behavior for existing templates. When unchecked, the subject renders verbatim in its source language everywhere it appears: the request list, request detail, agile board cards, notifications, and global search.

A request inherits the template's value at creation and carries it on the request. Later template edits do not change existing requests. The setting governs the subject only; note and description translation is unchanged, and the code-block workaround for notes still applies.

Requests: Duplicate RFC on Workflow and Project Links

The Duplicate Request action now appears on RFCs that are linked to a Workflow or a Project, not just standalone RFCs. This unblocks a common workflow: specialists routinely want to spin up a near-identical RFC that references an existing workflow or project, and the previous hard block forced them back to manual re-keying.

The action's visibility follows the same base eligibility as unlinked RFCs: request-creation permission, RFC category, no Case Management enabled, and specialist UI. No new manager-role gate is added. RFCs linked to a Problem remain fully blocked for all users, since problems sit on the incident side and are outside the scope of change management.

The duplicate itself is unchanged from the base feature. It retains the source's request template, so on save it instantiates a new workflow from that request template's workflow template, exactly as an original request created from the same template would. The duplicate is not attached to the source's workflow, project, or problem; it is a fresh, standalone RFC.

Currencies: Serbian Dinar (RSD) Added

Serbian Dinar (RSD) is now a selectable currency anywhere a currency can be chosen in Xurrent, including the Amount field on invoices, time entries, and expenses. Costs recorded in RSD convert to and from other currencies using the existing daily exchange-rate feed, and cost columns can be sorted and totaled in RSD like every other currency.

No existing data changes; the new option simply becomes available.