1. Real paid tasks → Demand Feature
We collect tasks from compliant sources, generate summaries, remove PII, and cluster similar tasks into demand features. Source text is never reproduced in full; pages retain summaries and outbound source links.
This is a traceable methodology, not a promise of business outcomes.
We collect tasks from compliant sources, generate summaries, remove PII, and cluster similar tasks into demand features. Source text is never reproduced in full; pages retain summaries and outbound source links.
FeatureStat records mention count, median budget, trend percentage, and regional breakdown. The trend board supports tech-stack, category, region, and time-window filters.
OpportunityAssessment stores heat, momentum, competition, risk, and a risk list. V1 uses heat 35%, momentum 25%, inverted competition 25%, and inverted risk 15%; it is a research-ranking tool, not a guarantee.
Development briefs, competition assessments, and opportunity risks are generated per cluster in batches by management commands and systemd timers, not by per-click LLM calls.
You can start evidence-bound real-time reports on the validation and comparison pages. Both actions use tokens; see Pricing。
Public pages provide summaries only. Unauthorized visitors never receive Pro body content: the detail view separates development_brief_verdict from development_brief and passes the latter only to entitled users.