Changelog

Added: Import Entire Websites into Your Knowledge Base

You can now add a full website to your Knowledge Base using a single root URL. The system automatically discovers pages and lets you control which ones are included using rule-based filtering.

How It Works

We added website-level crawling to the Knowledge Base inside Whippy.

Instead of adding individual URLs one by one, you now:

1. Enter a root domain or starting URL

2. The system automatically discovers pages across that site

3. You apply include and exclude rules to control which URLs are indexed

This update supports structured content ingestion, large documentation libraries, blog archives, help centers, and resource hubs.

Pattern-Based URL Filtering

You can filter discovered pages using path-based rules. This gives you granular control over which content enters your Knowledge Base.

Supported match types:

• Starts with
• Ends with
• Contains
• Exact match
• Custom (regex pattern)

Examples

Starts with /docs

Matches:

• /docs/getting-started
• /docs/install
• /docs/api/authentication

Contains help

Matches:

• /product/help/reset-password
• /support/help/account-access
• /help/contact-us

Ends with /archive

Matches:

• /blog/2022/archive
• /releases/v1/archive

Exact match /pricing

Matches only:

• /pricing

Custom rule /blog/.*/draft

Matches:

• /blog/2024/draft/new-feature
• /blog/october/draft/release-notes

Important Guidelines

• Website discovery is limited to 500 total pages per import to prevent over-crawling.
• Exclude rules always take precedence over include rules.
• Once published, a website import does not auto-refresh. New pages added to your site are not automatically re-indexed. Manual updates are required. Automatic refresh is planned for a future release.

Why It Matters

This update simplifies Knowledge Base management.

You no longer need to manually add individual URLs. You can import structured website content in bulk and control visibility with filtering rules. This improves:

• AI response accuracy
• Knowledge source consistency
• Content governance
• Documentation indexing
• Help center integration

Teams using Whippy for AI messaging, automated replies, and customer support workflows can now connect entire documentation portals or support sites in minutes.

If you are using Whippy’s AI-driven communication tools, this feature makes it easier to maintain an up-to-date, structured Knowledge Base that powers automated conversations.

Added: Conversation History Context for AI Agents

AI Agents now support optional access to prior conversation history, allowing them to continue conversations with more context.

How It Works

We added a new {{conversation_history}} variable that can be used in the AI Agent Editor. When this variable is included in an AI Agent’s instructions, the agent can reference past exchanges between the contact and the channel, such as SMS, chat or calls, before or during a live call or chat session.

This variable is available only in the AI Agent Editor and must be added manually. The feature is not enabled by default.

Why It Matters

This update allows AI Agents to respond with better continuity and awareness. Conversations can continue naturally when a contact switches from text to voice. Contacts no longer need to repeat information that was already shared earlier. AI responses feel more relevant because they are based on previous interactions, not just the current message or call.

This is especially useful for use cases like customer support, scheduling, lead qualification, and follow-ups where prior context improves accuracy and efficiency.

Important Notes and Behavior Changes

• Action required: Conversation history is not automatically included. If your use case depends on prior messages, you must add {{conversation_history}} to the AI Agent’s instructions.
• Changed behavior: AI Agents no longer receive conversation history by default. Previously, chat sessions included history automatically. This behavior was changed, and the variable is now required for all use cases where history is needed.
• UI updates: We added warning messages in the AI Agent Editor to indicate when conversation history is not included.

Added: Location-Specific VoIP Notification Preferences

Granular notification control per channel.

How It Works

We added channel-specific notification preferences that let users manage alerts on a per-channel basis. Notification settings now live on the relationship between a user and a channel, instead of being applied globally.

Previously, notification preferences were shared across all channels. Users could not mute one office without muting all of them. With this update, each channel (office, branch, or location) can be configured independently.

If no preference is set for a channel, the system defaults to the standard global settings.

Why It Matters

For multi-channel teams, notification overload reduces focus and productivity. This change gives users precise control over when and where they receive VoIP alerts, without affecting other active channels.

Teams can now stay responsive where it matters most and quiet where it doesn’t.

Key Benefits

• Reduced noise: Recruiters and managers only receive notifications from the channels they actively manage.
• Better user experience: Notification settings are no longer “all or nothing”.
• Scalable by design: As organizations add more channels, notification streams stay relevant and manageable.

Publisher Attribution for Knowledge Base Versions

Each knowledge base version now records who published it.

How It Works

Whippy now captures and displays the user who publishes each knowledge base version. When a version is pushed live, the system automatically records the publisher and associates that user with the published version. This happens at the time of publishing and requires no changes to existing workflows or permissions.

For versions published before this update, publisher details may not appear because that data was not previously tracked.

Why It Matters

Teams often collaborate on knowledge base content, with one user drafting and another approving or publishing. Previously, there was no reliable way to see who actually published a version. This update improves visibility and accountability by clearly showing who pushed each version live, which helps with audits, content reviews, and internal ownership.

This change supports better content governance across shared knowledge bases and aligns with Whippy’s focus on clear, trackable communication management.