What is question intelligence?
Question intelligence, or Q-IQ, is the demand-side signal of what the market is asking, captured in real time. It surfaces what people are confused about before that confusion shows up as volume or volatility.
The proprietary asset is the question stream
Market data is a commodity that many vendors sell. The stream of what people ask is not. It is a real-time map of information asymmetry: which names are drawing a cluster of questions, and which questions the engine cannot cleanly resolve.
Every product surface feeds this flywheel. Issuer Q&A, education, API explains, and the digital twin all record a question, anonymized, into aggregate signal.
From questions to a confusion signal
Aggregated, the question stream answers "what is the market confused about?" LyraMind computes an insights view (top symbols, top intents, sources, and a grounded-versus-refused ratio) and an Institutional Alpha Feed that ranks names by a confusion score. The score blends query volume with how unresolved those queries are, so a name drawing many questions the engine cannot ground scores highest.
The feed exposes where attention and unresolved questions cluster, never advice or a price prediction, and it stores no individual identity.
How LyraMind implements it
A best-effort recorder captures each question or query with its source, symbol, intent, and grounded flag, and never breaks an answer if recording fails. Aggregates track counts by symbol, intent, and source plus a grounding tally. The Alpha Feed reads over the recent window and ranks confusion clusters; it records nothing new and exposes no principal.
The current build is an in-memory MVP that starts the flywheel from launch. Persisting to Supabase is the productization step; the MVP proves the mechanism and surfaces the insight. The Alpha Feed is gated so an API key above the anonymous tier is required.
Related concepts
Question intelligence is fed by the trust envelope's grounded flag on every answer and by the predicted questions from the AI Readiness report. The Alpha Feed is its highest-margin expression.
- Q-IQ is the demand-side stream of what the market asks, anonymized
- Every surface feeds it: issuer Q&A, education, API explains, the digital twin
- Insights view: top symbols, intents, sources, and a grounded-versus-refused ratio
- The Alpha Feed ranks names by a confusion score blending volume and unresolved rate
- Records nothing new for the feed and stores no individual identity
- In-memory MVP today; Supabase persistence is the productization step
Questions
Does question intelligence track individuals?
No. The flywheel stores no principal. The insights and Alpha Feed are aggregate and anonymized.
What is a confusion cluster?
A name drawing a cluster of queries, scored by volume amplified by how unresolved those queries are. High scores flag structural confusion the engine cannot cleanly resolve.
Is the Alpha Feed a prediction?
No. It is an aggregated, anonymized signal of where attention and unresolved questions cluster. It is never advice or a price prediction.