Stop shipping SEO content you can't defend.
Source-backed briefs. Quality and GEO gates that block bad drafts. AI visibility tracking across ChatGPT, Perplexity, and Google AI Overviews. One operator surface — not a marketing site dressed as one.
What you see on day one
An operator dashboard built for the work, not the demo.
Live pipeline progress with elapsed time. 14-day throughput and completion-rate trends. One-click bulk retry on anything that failed. No marketing fluff between you and the run.
Articles this week
47
+18%
Running now
3
live
Completion score
94
+6%
GEO proxy
88%
+2%
Throughput · 14d
Status distribution
Completion rate · 14d
Running pipelines
3 livebest running shoes for flat feet
korean medical tourism 2026 guide
wholesale k-beauty distributor india
What it actually does
Four things, named and visible.
Source-backed briefs
SERP, entity coverage, and GEO context get pulled before any writing starts. Every claim in the draft maps to a citation you can defend.
Pulls live SERP + DataForSEO competitor data
Quality gates that block
Citability, entity coverage, GEO score, copywriting, technical SEO. The pipeline pauses for operator review instead of shipping garbage.
Five named score gates, all visible to the operator
AI visibility, named
Track citations across ChatGPT, Perplexity, and Google AI Overviews. See invisible gaps where competitors get cited and you don't.
Mentions, impressions, source-of-truth coverage
Multi-client, no leakage
Per-tenant routing. Per-tenant daily quotas. RLS policies you can audit. Client A never sees Client B's threads, by URL or otherwise.
Tested across two tenants on every deploy
How a single article moves through the system
Five phases. Operator review between each.
- Phase 11
Ingest
Keyword, target URL, audience. Brief skeleton populated.
- Phase 22
Diagnose
SERP scrape, competitor gap, entity coverage map.
- Phase 33
Strategize
Angle decisions, GEO targets, citation budget.
- Phase 44
Architect
Outline with named entities and source links per H2.
- Phase 55
Create
Draft → score gates → operator approval → publish.
Each phase pauses for review if a gate fails · No silent shipping
Beta access posture
Production hardening, not security theater.
Six things were already true on day one of the closed beta. They're checked again on every deploy via scripts/smoke_test.sh before any beta invite goes out.
- BYOK Fernet encryption for client API keys
- Single-origin app shell · API runs internal-only
- 5xx responses sanitized · no Python traceback leaks
- Per-tenant daily run cap · Redis-backed quota
- Sentry-wired backend + frontend · request_id surfaced
- Tenant isolation tested across two users on every deploy
Ready to run a pipeline you can actually defend?
Beta seats are limited while we tune the per-tenant cost ceiling. Existing operators can sign in below.