Mentionwell measures how answer engines describe and recommend a brand, identifies the actual evidence gap, publishes the intervention that gap supports, and then remeasures the same prompts.
Build an autonomous publishing system that treats citations, recommendations, and search outcomes as different things - and can prove what changed after it publishes.
Role
Product owner and operator across the system concept, automated pipeline, editorial standard, and commercial direction.
What It Does
Postgres · Multi-Model Pipeline · 10 AI Engines Measured
01
Ten-Engine Citation Tracking
Runs the same prompt across ChatGPT, Claude, Perplexity, Gemini, Grok, Google AI, and more - in parallel - and records who cited the brand.
02
Agentic Onboarding
Point it at a domain: it crawls the site, classifies the CMS, and confirms tone and topics in one guided session.
03
Full Editorial Pipeline
Research, outline, parallel section writing, a critic pass, humanizing, imagery, and publish - every stage a tunable prompt.
04
Crawl-to-Citation Proof
Logs when AI crawlers read a published article and ties that crawl to the citations that follow.
05
Share of Voice
Measures the brand's slice of AI answers over time, refreshed every hour.
06
Publishes Anywhere
A reader API, signed webhooks, GitHub MDX, a WordPress plugin, an npm CLI, and MCP servers.
Real ProductMentionwellResearch · Publish · Measure
