01
Review inbox
The work queue combines App Store and Google Play reviews with filters, status, priority, due dates, assignees and labels.
For app teams that answer to users
Turn app reviews into replies, decisions and roadmap.
ReviewRelay brings App Store and Google Play reviews and customer feedback into one self-hosted workspace. AI writes the drafts, your team approves them, and the insights land straight with product.
Self-hosted · Multi-tenant · German and English
Connects with
- Apple App Store
- Google Play
- Slack
- Microsoft Teams
- Telegram
- GitHub
- OpenAI
- REST API and MCP
Features
Everything between a review and a reply.
From intake through draft and approval to analysis: one workspace instead of many tabs.
One inbox for every store
Reviews from the App Store and Google Play in one place: saved views, bulk actions, assignments and an Inbox Zero mode.
AI drafts in your brand voice
Reply drafts in the reviewer's language, with brand rules, an explanation, and learning from the replies you approve.
Quality gate and approvals
Rules decide what needs sign-off. A quality check scores every draft before it is published.
Customer feedback beyond the stores
Collect and triage feedback, and turn it into feature requests and follow-ups so nothing gets lost in an inbox.
Insights and trends
Trend detection, a knowledge graph and an analyst you can ask questions about your reviews. Evidence can be shared by link.
Incidents and severity
Recurring problems are grouped into incidents and prioritised by severity and SLA. Abuse and spam are detected.
Notifications and hand-offs
Slack, Microsoft Teams, Telegram, email and push. Hand-off links get support cases to the right person without an account.
API and MCP
A REST API at /api/v1 and an MCP endpoint, so your own tools and AI assistants can work with your reviews.
How the team works
From the first star to the insight.
- 1
Connect
Link App Store Connect and Google Play. New reviews arrive automatically.
- 2
Understand
Translation, classification, labels and severity are ready when someone opens the review.
- 3
Reply
Check the draft, pass the quality gate, approve, publish.
- 4
Learn
Trends, incidents and feature requests show what the product needs next.
Screens
What a working day looks like.
The real interface with demo data, in light and dark mode.
02
Review with AI draft
The review detail shows the rating with its metadata next to the response composer and an AI draft that passed the check and awaits approval.
03
Approvals
Submitted responses appear next to the original review, with the quality check result, overdue state and actions to revise or approve.
04
Analytics
Analytics shows key metrics with trends against the previous period, signals that need attention and the start of the volume and rating distribution charts.
05
Issues and incidents
Issue clusters built from negative reviews with severity, suspected cause, early signal and confidence, plus an active outage mode banner.
06
Feature requests
Feature requests grouped from reviews are ranked by demand, with status, growth, territories and versions.
07
Analyst
The analyst answers plain-language questions about the review data and shows the product health score per app below.
Security and operations
Your data stays with you.
ReviewRelay runs on your own server. You decide what the AI gets to see and what it may cost.
Separated tenants
Every tenant has its own PostgreSQL schema. Policies check every access.
Modern sign-in
Passkeys, TOTP and OIDC. Passwords hashed with Argon2id, sensitive actions need a fresh confirmation.
Encrypted secrets
Credentials are encrypted with your own master key. Backups are encrypted and restored as a weekly drill.
AI with privacy and budget
Privacy settings define what goes to the AI provider. Usage and cost are visible and can be capped.
One container plus a database
Elixir and Phoenix with PostgreSQL 17. A single Docker Compose command gets you started.
Protected outbound traffic
Outgoing requests go through one central layer with protection against SSRF.
FAQ
Frequently asked questions
Which stores are supported?
The Apple App Store through App Store Connect, and Google Play. Reviews in any language are translated and classified.
Does the AI reply without human review?
Only as far as you allow it. Drafts go through the quality gate and the approval rules you define yourself.
Where is the data stored?
In your own PostgreSQL database on your own server. Only what your privacy settings allow is sent to the AI provider you configure.
How do I get started?
With Docker Compose: one app and PostgreSQL 17. On first visit the setup wizard creates the first administrator, after which you can add tenants and apps.
Is the interface available in English?
Yes. The whole interface comes in German and English, each in light and dark mode.