ShipPulse
  • Pricing
  • Docs
  • Blog
  • Compare

Product

TestimonialsChangelogStatus PagesFeedbackRoadmapPricing

Resources

DocsBlogAPI ReferenceSDKHelp Center

Company

AboutContact

Legal

TermsPrivacyCookie PolicyDPASub-processors

Product updates

Changelog updates only. Unsubscribe any time.

ShipPulse operated by Igor Bogdanov, Limassol, Cyprus. [email protected]. Cyprus registration number pending — will be published once issued.

ShipPulse

© 2026 ShipPulse. All rights reserved.

OverviewQuick StartCore ConceptsWidget ReferencePlatform GuidesAPI ReferenceAPI PlaygroundError HandlingPaginationRate LimitingJavaScript SDKWebhooksZapier, n8n & MakeCustom DomainsTeam ManagementBilling & PlansNotification ChannelsAI FeaturesOverviewFrom SenjaFrom Testimonial.toFrom HeadwayFrom Canny

Introduction

  • Overview
  • Quick Start
  • Core Concepts

Embed Widgets

  • Widget Reference
  • Platform Guides

REST API

  • API Reference
  • API Playground
  • Error Handling
  • Pagination
  • Rate Limiting

SDK & Webhooks

  • JavaScript SDK
  • Webhooks
  • Zapier, n8n & Make

Guides

  • Custom Domains
  • Team Management
  • Billing & Plans
  • Notification Channels
  • AI Features

Migrations

  • Overview
  • From Senja
  • From Testimonial.to
  • From Headway
  • From Canny

Need help?

[email protected]
Configuration

AI Features

ShipPulse uses large language models to automatically analyze testimonials and feedback, surface the most useful content, and reduce manual moderation work.

AI processing happens asynchronously in the background — it doesn't block submission or display of content. Available on Pro and Agency plans.

Sentiment analysis
Highlight extraction
Duplicate detection
AI changelog summaries

Sentiment analysis

Every testimonial is automatically scored for sentiment when it is submitted or imported. Sentiment analysis runs on the full text of the testimonial, not just keywords.

Sentiment scores

LabelScore rangeMeaning
positive0.67 – 1.0Clearly positive feedback, suitable for featuring
neutral0.34 – 0.66Mixed or factual feedback without strong polarity
negative0.0 – 0.33Negative feedback — flag for review

Where to see it

In your dashboard, every testimonial card shows a sentiment badge (Positive / Neutral / Negative). The Testimonials list can be filtered by sentiment. High-volume projects will also see a sentiment distribution chart in Analytics.

Highlight extraction

For each testimonial, the AI extracts the single most impactful sentence or phrase — the "highlight". This is shown as a callout on the testimonial card and used in the Wall of Love widget to show the most compelling quote.

How it works

1The full testimonial text is analyzed by the language model.
2Sentences are scored based on specificity, emotional impact, and relevance to product benefits.
3The highest-scoring sentence is stored as the highlight. If no sentence exceeds the threshold, no highlight is set.
4You can override the AI highlight by manually selecting a different sentence in the testimonial editor.

Where to see it

Highlights appear as a styled pull-quote on testimonial cards. In Wall of Love and Carousel widgets, the highlight is shown prominently below the star rating. Highlighted testimonials also get priority in "Best" sort mode.

Duplicate detection

When a user submits a bug report or feature request on your Feedback Board, ShipPulse checks for semantically similar existing submissions — not just exact keyword matches.

Algorithm

1The new submission is embedded using a vector embedding model.
2The embedding is compared against the embeddings of all open feedback items using cosine similarity.
3Items with similarity above the threshold (default: 0.85) are flagged as potential duplicates.
4The submitter sees a 'Similar requests' list and can upvote an existing item instead of creating a new one.
5You can merge flagged duplicates from the moderation queue — votes are combined into the canonical item.

Where to see it

The Feedback moderation queue shows a "Possible duplicates" indicator on flagged submissions. You can dismiss the duplicate suggestion or merge items with one click.

AI changelog summaries

When creating a changelog entry, ShipPulse can generate a customer-friendly summary from your technical release notes or Git diff. The AI rewrites jargon into plain language suitable for non-technical users.

How to use it

1. Open the changelog editor and click ✨ Generate summary.

2. Paste your technical notes or leave the body field with raw text.

3. The AI generates a customer-friendly version. Review and edit before publishing.

4. The original technical text is preserved as an internal note if needed.

AI summaries are generated by OpenAI GPT-4o. Content is not used to train models. See our Privacy Policy for data handling details.