v5.15 — July 2026

Changes since v5.14 (December 2025). This release spans the platform, connectors, agents, and the admin experience. Below is a customer-facing summary; internal refactors and infrastructure work are omitted.

• New Connectors & Integrations
• Branding & White-Labeling
• AI & Assistant Data Access
• Configuration Builder Experience
• Data Sync & Mapping Engine
• Webhooks & External Access
• Reliability & Performance
• Users & Access Management
• Documentation

New Connectors & Integrations

We expanded the connector catalog across HR, communications, logistics, procurement, accounting, CPQ and more — and made existing connectors more capable and easier to authenticate.

New connectors: Slack (messaging & alert delivery), Paycor (HRIS), Paycom (HRIS), Gaiia (billing), Talkdesk (contact center), Track Unit (telematics), Wherefour (ERP/traceability), Fireflies.ai (meeting intelligence), Salesmate (CRM), Tai (TMS), Cincom CPQ, SAP Ariba (cXML procurement), Sage Intacct (accounting/ERP), Loopio (RFP/proposals), FedaDB (database), and PCI Pal (secure payments).

• Multi-auth connectors — Connectors can now offer more than one authentication method, so you pick the one that fits your environment.
• Simpler, more reliable authentication — API-orchestrated OAuth with PKCE, automatic detection of expired refresh tokens with a one-click re-authenticate action, and connection verification built into every connector's Test action.
• Deeper existing-connector coverage — Extended Eptura Asset, NetSuite 2026.1 support, Egnyte moved to long-lived refresh-token OAuth, Rental Result SOAP handling, and corrected Mortgage Automator & WInTeam definitions.
• File System as a connector — Local and network file locations are now a first-class connector you can configure like any other system.

Branding & White-Labeling

A complete white-labeling system lets partners and enterprises present the platform under their own brand — no code required.

• Full brand kit — Upload your logo, assistant avatar, page title and web manifest; set your primary color and theme; override favicon, Apple touch icon and Windows tile.
• Light & dark variants — Provide distinct brand assets for light and dark mode.
• Effortless editing — Drag an image onto the preview to upload it, watch changes render live, and see updates propagate instantly to other open tabs.
• Branding at provisioning — New customer environments can be stood up with default branding already applied.

AI & Assistant Data Access

Your synced data is now directly usable by AI assistants (via MCP) — for retrieval, analysis and data entry — with the guardrails and performance to use it at scale.

• Rich queries — Filter records, compare one property against another, and use StartsWith/EndsWith operators.
• Follow the relationships — Expand related records inline (joins), filter on those related properties, and shape the response to only the fields you need.
• Aggregations — Sum, Average, Count, Min, Max and Distinct-Count with group-by, including across joined relationships.
• Fast lookups — Bulk get by ID list, scalable keyword search with paging, and smart routing of ID lookups.
• Conversational data entry — Guided forms let the assistant create and update records, with fields scoped by transaction type.
• Saved queries — List and run saved queries directly from the assistant.
• Governance — Per-tool permissions, query logging and result counts keep AI access controlled and auditable.

Configuration Builder Experience

The admin experience for building and testing integrations got a major usability pass — faster feedback, fewer surprises, and clearer editing throughout.

• Map Preview — Test map groups in real time from the mapping editor and see exactly what output a configuration produces before you deploy.
• Drag & drop mapping — Reorder and restructure maps by dragging, with a unified Properties tab that brings related settings together.
• Inline config validation — Configuration errors are surfaced right in the forms as you build.
• Better Test Connector — Test read actions from the data-source tab, run SQL tests with copy-able result cells, clearer loading and failure feedback, and reliable error toasts.
• Safer edits — Type-to-confirm before deleting high-impact configurations, plus a "Compare with Live" action.
• Everyday conveniences — Timezone typeahead, CSV template generator, "Copy Webhook URL", extra primary keys in Maintain Index, and editors for connector- and builder-level AI instructions.

Data Sync & Mapping Engine

The core engine gained substantial new mapping power — more transforms, smarter parent/child handling, and finer control over how records are matched, filtered and written.

• New transforms — Convert to UTC, US state name ↔ abbreviation, and IANA timezone normalization.
• Parent/child mapping — Map from a parent entity (including multi-hop ancestors), skip parent saves while processing children, and skip inner saves when children fail.
• Looping lists — Conditional read/write filters, delete-aware maps that remove embedded child items, and in-place grouping.
• Record matching — Chained existence lookups, key-extraction actions on save, subscriber-side primary-key selection, and extra-key lookups.
• Parallel publishing — Publish work concurrently for higher throughput.
• Session values — Capture and reuse values across a run, including across multiple lists.
• Flexible paging — Cursor-based and "has-more" paging strategies for large data sets.
• Queue publishers — Queue-message and queue-release publishers, with release-if-trigger-arrives-first handling.
• More control — Variables in data selectors, insert-at-root for XML, and trimming preload responses to just the fields you need.

Webhooks & External Access

Real-time inbound events and outbound lookup endpoints make it easier to connect the platform to the tools around it.

• Inbound webhooks — A dedicated webhook endpoint accepts events from external systems, with a one-click "Copy Webhook URL" and per-connector signature verification.
• Cross-reference lookups — A dedicated cross-reference endpoint, plus a Power Automate custom connector, lets other tools resolve entity IDs across systems.
• Event fan-out — A single inbound document can drive multiple downstream records via response-selector fan-out.
• API Gateway — Create and manage API gateways directly from the admin experience.
• Clearer API errors — Platform API responses now return human-readable, actionable error messages.

Reliability & Performance

A large share of this release went into making sync more dependable and operations more hands-off.

• Self-updating agents — Remote, secure agent updates with identity certificates from a managed vault, a guided Windows installer, and automatic upgrade of older agents.
• Messaging hardening — Long-running saves no longer lose their message lock, redeliveries settle correctly, and obsolete subscriptions are cleaned up automatically.
• Resilient outbound calls — Idempotency-aware transient retry on outbound API requests recovers from momentary network and vendor hiccups without creating duplicates.
• No more stuck runs — Data pulls that hang now time out cleanly instead of taking down processing, and dependent-publisher runs report status correctly.
• Better error visibility — Failures to record message metadata are now surfaced instead of silently lost.
• Search & bulk processing — Optimized search-entity sync and bulk processing for large data volumes.

Users & Access Management

• User provisioning API — Create users programmatically as part of your onboarding workflows.
• Manage users at scale — A user grid with mass-edit functions and role deletes that cascade cleanly to affected users.
• Tool-level permissions — Fine-grained permissions govern which AI tools each user can access.

Documentation

• Web-based User Guide — This guide: a fast, searchable site with cross-linked topics and navigation.
• Mobile friendly — Improved table and layout treatment for reading on any device.
• Authentication reference — Supported authentication methods are now fully documented.