Whelp

Aggregates incoming support inquiries from email, chat, social media, help desk, and other channels into a single unified inbox interface, using channel-specific connectors that normalize message metadata (sender, timestamp, channel origin) into a common data model. Messages are threaded by conversation context rather than channel, allowing agents to view full customer interaction history across platforms without switching tabs or losing context.

Generates contextually relevant draft responses to customer inquiries using a pre-trained language model (likely GPT-3.5 or similar), triggered when an agent opens a ticket. The system analyzes the customer message, channel context, and (optionally) previous conversation history to produce 1-3 suggested reply options that agents can accept, edit, or reject. No fine-tuning or custom training data is required, enabling immediate deployment without knowledge base setup.

Automatically converts incoming emails into support tickets, parsing sender information, subject, and body content into structured ticket fields. The system likely uses email forwarding or IMAP integration to capture emails, extracts key information (customer name, email, issue description), and creates a ticket in the unified inbox. Attachments may be preserved and linked to the ticket.

Routes incoming support messages to appropriate agents or teams based on channel origin, message content, or predefined rules. The system likely uses simple rule-based routing (e.g., 'all Instagram DMs go to Team A') rather than ML-based intelligent routing, and assigns tickets to available agents with load-balancing to prevent bottlenecks. Routing rules are configurable via UI without requiring code.

Builds a unified customer profile that aggregates all interactions across connected channels, displaying conversation history, contact information, and engagement metadata in a single view. The system likely uses email address or phone number as the primary identifier to link messages from different channels to the same customer, and maintains a timeline of all interactions regardless of channel origin.

Automatically generates concise summaries of support tickets and assigns category/topic tags using NLP classification. The system likely uses pre-trained models to extract key information from customer messages and conversation history, producing summaries that help agents quickly understand ticket context and enabling filtering/search by category. Categorization may be rule-based or ML-based, but appears to use predefined categories rather than custom taxonomy.

Enables support agents to collaborate on tickets through internal notes, @mentions, and team communication without exposing internal discussion to customers. The system likely uses a comment/note thread attached to each ticket, with notifications triggered by @mentions, allowing agents to request help, share context, or escalate issues without creating separate communication channels.

Offers a free tier with limited features (likely basic inbox consolidation, limited AI suggestions, small team size) and paid tiers that unlock advanced features (more AI suggestions, advanced routing, analytics). The freemium model is designed to allow bootstrapped teams to start without cost, with clear upgrade paths as they scale. Pricing tiers appear to be based on team size, message volume, or feature access rather than per-agent seats.

Formats outgoing agent responses appropriately for each channel (email, chat, social media) and delivers them through the correct channel API, handling platform-specific constraints (character limits, formatting rules, media support). The system likely maintains a mapping of response format to channel type, automatically converting rich text or markdown to platform-native formatting (e.g., Slack formatting for Slack, plain text for email).

Provides a dashboard displaying key support metrics such as ticket volume, average response time, resolution time, and agent performance. The system likely aggregates data from all connected channels and presents it in simple charts/tables, enabling teams to track support efficiency and identify bottlenecks. Analytics appear to be basic (no advanced segmentation or predictive analytics) and may be limited to freemium tier.

Allows teams to define custom ticket states (e.g., 'New', 'In Progress', 'Waiting for Customer', 'Resolved') and move tickets through a workflow pipeline. The system likely provides a kanban-style board or list view showing tickets grouped by status, with drag-and-drop or button-based state transitions. Workflow automation may trigger actions on state changes (e.g., send follow-up email when ticket moves to 'Waiting for Customer').