Changelog
Release notes
A record of everything we've shipped — new features, improvements, and fixes.
Initial Launch
Noma is live. After months of beta testing with early users, we're shipping version 1.0.0 with the full feature set: everything you need to track, audit, understand, and improve your brand's presence in AI- generated answers.
What's included
- Visibility Tracking across 5 AI engines. Run your brand against a curated set of prompts across ChatGPT, Perplexity, Gemini, Claude, and Grok. See your mention rate, position, and description accuracy for each engine — updated weekly.
- Site Health audit with 12 signals. Noma crawls your domain and checks for the structural signals that AI engines care about: Organization schema, FAQ schema, llms.txt, Open Graph metadata, canonical URLs, page speed, and more. Each signal is scored and prioritized so you know exactly what to fix first.
- Gap Research. Identify the topics and questions where your competitors are appearing in AI answers and you aren't. Gap Research surfaces the exact prompts that return your rivals — and gives you a content brief to close the gap.
- Content Generation. Generate AEO-optimized content briefs and full articles that are structured for AI engine inclusion — not just keyword density. Every piece of content includes recommended FAQ schema and structured data annotations.
- Sage — your in-product AI strategist. Sage is Noma's conversational AI layer. Ask it why you're not appearing in a specific engine, what your highest-priority fix is, or how a competitor is outperforming you. Sage has full context of your account data and gives specific, actionable answers.
Thank you to everyone who tested Noma during the alpha and beta phases. Your feedback shaped every feature on this list. We're just getting started.
Beta — Core tracking and audit engine
Beta is open. Invited users can now access the core tracking and audit features that form the foundation of Noma. This release is focused on the two workflows we validated most heavily in alpha: knowing where your brand stands in AI answers, and understanding why.
What's new in 0.9.0
- AI visibility tracking for ChatGPT and Perplexity. Run a curated prompt set against both engines and see your mention rate, position in answers, and a log of the raw AI responses. Gemini, Claude, and Grok support are coming in 1.0.
- Basic site health audit. The audit engine checks eight of the twelve signals planned for 1.0: Organization schema, FAQ schema, llms.txt, robots.txt, Open Graph metadata, page title length, canonical URLs, and HTTPS. Each check includes a pass/fail status and a short explanation of why it matters for AEO.
- Foundational dashboard. A single-page overview showing your current visibility score, audit status, and a prioritized list of recommended actions. The dashboard updates automatically after each tracking run.
Known limitations in this release
- Tracking is limited to ChatGPT and Perplexity. Additional engines in 1.0.
- Prompt scheduling runs weekly. Manually-triggered runs are not yet available — coming soon.
- Content generation and Gap Research are not included in this beta release.
- Sage (the AI strategist) is not available in beta. It ships with 1.0.
If you're a beta user, your feedback in the next few weeks will directly shape the 1.0 feature set. Use the in-app feedback button or email us directly.
Alpha — Foundation
The infrastructure is live. Version 0.8.0 is an alpha release for internal testing and a small group of design partners. No public access yet — this release is about validating the foundation before we build the product layer on top of it.
What's live in 0.8.0
- Supabase authentication. Email/password sign-in and magic link support. Session management, token refresh, and server-side auth via Next.js middleware are all working. Row-level security is enabled on all tables.
- Multi-organization support. A single user account can belong to multiple organizations, each with its own settings, projects, and billing. Organization switching works from the sidebar.
- Project creation. Users can create projects tied to a domain. Each project has its own prompt configuration, audit history, and visibility data. The project model is designed to support agency use cases with multiple client domains.
- Basic prompt management. A list of seed prompts per project category. Design partners can add custom prompts and tag them by intent (awareness, comparison, purchase). The prompt runner is implemented but not yet connected to the live AI engines in this build.
The focus for the next six weeks is connecting the prompt runner to real AI engine APIs, building the audit crawler, and shipping the first version of the visibility dashboard. If you're a design partner and you have feedback on the project setup flow, now is the time to share it.