Feature Request: Sentiment Analysis on Inbox automatically classify incoming comments and messages by emotional tone

[githb-ac]

TryPost's unified inbox brings comments, mentions, and messages from all connected platforms into one place but as inbox volume grows, the challenge shifts from consolidation to prioritisation. A brand receiving hundreds of comments a day cannot manually read every one to identify which are urgent complaints, which are enthusiastic advocates worth engaging, and which are neutral or irrelevant. This feature request proposes native sentiment analysis on the TryPost inbox that automatically classifies every incoming message and comment by emotional tone, allowing community managers to triage, prioritise, and respond to the right conversations first without reading everything manually.

Summary

An AI-powered sentiment classification layer applied to every incoming comment, mention, and direct message in the TryPost unified inbox. Each item is automatically tagged as positive, neutral, negative, or a more granular category such as complaint, question, praise, or crisis signal. Users can filter the inbox by sentiment, set up alerts for negative sentiment spikes, and view sentiment trends across their accounts over time turning the inbox from a raw message list into an intelligent engagement prioritisation system.

Why This Matters

For any brand managing an active social presence, missing a public complaint or a viral negative comment thread is a reputational risk. Community managers currently have no systematic way to surface high-priority negative sentiment from within a high-volume inbox without reading everything which does not scale. Sentiment analysis solves this at the infrastructure level by ensuring that the most urgent conversations are always visible regardless of inbox volume. It also produces aggregate data that is genuinely valuable to brand and marketing teams knowing that 30% of this week's comments on a product launch post expressed confusion is a signal that the messaging needs to be addressed, and that insight currently requires manual analysis that almost no team has the capacity to do.

Proposed MVP

Automatic sentiment classification on every incoming comment, mention, and DM positive, neutral, negative

Granular sub-tags for common intent categories including complaint, question, praise, feature request, and spam

Inbox filter by sentiment so community managers can view all negative comments across all platforms in a single prioritised view

Sentiment spike alert a notification triggered when negative sentiment volume on a specific post or account exceeds a defined threshold within a short time window, indicating a potential crisis or viral complaint thread

Sentiment trend dashboard showing the proportion of positive, neutral, and negative sentiment across all inbox activity over a selected date range, broken down by platform and post

Sentiment score per post visible in the analytics dashboard showing the overall emotional response a piece of content generated from its comments

Manual override allowing community managers to correct a misclassified sentiment tag, which feeds back into improving classification accuracy over time