The full Listen → Decide → Build framework is live. 6 channel types, evidence-based scoring, 23-tool AI co-pilot, competitive intelligence, and built-in Tracker. This is MindBacklog v1.
- NewFeedback Widget — Embeddable JavaScript button for any website or web app. Customizable accent color, position, trigger label, and categories. User identification via
identify API. Domain restrictions for security.One line of code. Real-time AI processing on every submission.
- NewAsk Mind Co-Pilot — 23 tools — 12 read-only + 11 mutation tools, all via natural language. Query signals, create features, write PRDs, reprioritize your backlog, assign owners and teams, schedule dates. Confirm-before-execute for all mutations.Not a chatbot — a product management co-pilot with deep product memory.
- NewCompetitive Intelligence — Monitor competitor app reviews (Apple + Google Play) and competitor websites. All competitive signals run through the same AI pipeline with a market_signal lens. Ask Mind answers competitor questions.Key differentiator: no other PM tool monitors your competitors automatically.
- NewTracker — Built-in Execution Board — Kanban workflow for user stories and bugs within features: Draft → Ready → In Progress → Ready for QA → In QA → Waiting Acceptance → Done. Story points, severity, priority, sprint tracking. Threaded comments and activity logs.
- NewAI Content Generation — PRDs generated from product context + linked customer signals. User stories with acceptance criteria. Problem statements, business value, acceptance criteria — individual artifact fields generated on demand.
- NewPublic Portal — Public-facing feature request page for customers. Browse existing requests and submit new ones. Portal channels auto-managed through Product Settings.
- ImprovedApp Review Pipeline v2 — Reviews now process through the full 4-phase AI pipeline: Enrich → Validate (Bouncer) → Classify (Sorter) → Route (Inbox Attach). Skips already-fetched reviews. AI summaries generated per review.
- ImprovedChannel editing & management — All channel types now support inline editing of name, weight, sync frequency, and configuration. Unified settings panel across channel types.
- ImprovedIn-app notifications — Get notified when signals are linked, features are updated, and comments are posted. No self-notifications — only actions by others trigger alerts.
- FixDrawer comment submission — Comments in feature, user story, and bug drawers now submit via AJAX without page refresh. Comment count updates in real-time.
- InfraPaddle billing integration — Per-seat pricing across 3 tiers. 14-day free trial. Founding member pricing at 50% off while subscribed. Webhook-driven subscription lifecycle management.
Major upgrade to the product intelligence layer — multi-level summaries, smarter relevance filtering, and faster context retrieval.
- NewMulti-level summary strategy — Master summaries, source summaries, and chunk embeddings now work together for richer product context with fewer tokens.Reduces AI cost per query by ~40% while improving context accuracy.
- NewSource relevance validation — MIND now validates whether ingested content is actually about your product before including it in context. Competitor mentions and generic industry content are filtered out automatically.
- ImprovedProduct master summary regeneration — Summaries now update incrementally as new sources are added instead of regenerating from scratch. Faster and more cost-efficient.
- Infrapgvector embedding storage — Migrated vector storage to PostgreSQL pgvector extension for better query performance and simplified infrastructure.Removed dependency on external vector DB service.
- InfraRedis job queue optimization — Background processing pipeline now handles source ingestion 3x faster with improved retry logic and dead letter handling.
AI-powered prioritization scoring is live. Every feature in your Ideas bucket now gets automatic WSJF and RICE scores grounded in linked customer signals.
- NewWSJF scoring engine — Calculates Cost of Delay, User/Business Value, Time Criticality, and Risk Reduction using your product context. Scores update as MIND learns more.
- NewRICE scoring framework — Reach, Impact, Confidence, Effort calculated per feature. Toggle between WSJF and RICE views in the Ideas bucket.
- NewRevenue impact estimation — AI estimates potential MRR impact based on feature type, user segment, and product context.Accuracy improves as you add more product context to MIND.
- ImprovedFeature detail panel — Scores, evidence sources, and AI-generated descriptions now show in a unified side panel with tabbed navigation.
- DesignScore visualization bars — Horizontal score bars with color-coded severity. High-impact features glow. Scannable at a glance.
The core intelligence feature: raw signals from all sources are now automatically clustered into distinct, actionable features in the Radar.
- NewAutomatic signal clustering — AI groups “search is slow,” “I hate the search,” and “need better filters” into one feature — with links to every original source.
- NewCluster confidence scoring — Each cluster shows a confidence percentage. Low-confidence clusters are flagged for manual review.
- NewManual merge & split — Disagree with the AI? Merge clusters together or split them apart. MIND learns from your corrections.
- ImprovedRadar feed sorting — Raw signals now sortable by source, date, sentiment, and processing status.
- FixEmail parsing edge cases — Fixed issue where forwarded emails with nested reply chains were being treated as multiple separate items.
Connect your feedback sources and watch everything flow into a single feed. The Listen layer is live.
- NewApp Store review ingestion — Connect your Apple App Store app. Reviews sync automatically with sentiment tagging.
- NewGoogle Play Store ingestion — Same as App Store — ratings, reviews, and sentiment from Google Play.
- PlannedReddit monitoring — Track specific subreddits. Threads mentioning your product will be ingested with full context. Coming soon.
- NewEmail ingestion — Unique inbound email addresses (e.g., acme-support@mindbacklog.net). Real-time IMAP processing.
- NewCustomer web forms — Generate unique form links. Submissions flow directly into the Radar.
- DesignUnified Radar feed UI — All sources in one scrollable feed with source badges, timestamps, and sentiment indicators.
- InfraBackground processing pipeline — Laravel + Redis queue system for async source polling and ingestion. Handles rate limits gracefully.
The starting point. Authentication, product creation, and the bones of the three-bucket architecture.
- NewAuthentication system — Email + password and Google OAuth. Multi-organization support with role-based access control.
- NewProduct workspace creation — Create a product, upload docs or provide a URL, answer clarifying questions. MIND builds initial context.
- NewThree-bucket architecture — Radar → Ideas → Roadmap. Features move between buckets with status-specific workflows at each stage.
- DesignDark command-center UI — Midnight blue backgrounds, neon cyan accents, JetBrains Mono data typography. Designed for PMs who live in dark mode.
- InfraCore stack deployed — Laravel backend, PostgreSQL with pgvector, Python AI services, Redis queues, Docker containerization. CI/CD pipeline operational.Supabase S3-compatible storage for production files.
MindBacklog