Research Intelligence
Evidence · market reality · governance · knowledge
Operator access to real research evidence, market-data integrity, governed decisions and versioned knowledge.

Market data · research validation · governance · knowledge · autonomous specialists.
Evidence · market reality · governance · knowledge
QNChoose a real symbol, provider, venue and resolution. QuantNova will not draw synthetic history.
Real canonical data from Telegram, Portal, Cloud Storage and Research Brain.
Its actual pipeline, evidence and runtime diagnostics will appear here.
—
The browser session authorizes the real production Research Portal. No API key is exposed to JavaScript.
Real diagnostics, verified runtime agents, research-platform health and governed remediation boundaries. No synthetic metrics.
| Agent | Role | Mode | Status | Tasks | Orchestration OK | Validation | Last error |
|---|
Live direct-exchange market intelligence with measured price/time, trades, L1 BBO, L2 depth/DOM and credential-gated real L3 order events. Every displayed level exposes its provider, freshness, latency and actual availability.
| Side | Price | Size | Order ID | Updated |
|---|
| Provider | Venue | Symbol | Level | Events | Gaps | Mean latency | Last event | State |
|---|
| Provider | Symbol | Level | State | Sequence | Gaps | Out of order | Duplicates | Dropped | Reconnect / Resync | Latency | Clock skew | Feed age |
|---|
| Provider | Asset | Level | Technical | Credential | Entitled | State | Universe |
|---|
Observed news, structured events, geopolitical coordinates, research evidence, governance votes, incidents and point-in-time system state. Correlation and coverage similarity are never presented as causation.
| Cluster | Sources | Coverage | Claim disagreement |
|---|
| Decision | Votes | Uncertainty | Risk | Outcome | Updated |
|---|
| Severity | Incident | Status | Opened | Recovery |
|---|
Choose a historical point in time.
Research claims become preregistered hypotheses before market outcomes are inspected. Point-in-time validation, holdout testing, multiple-testing control and replay prevent research from becoming untested belief.
| Hypothesis | Agent | Feature | Data | Horizon | Sample floor | Status | Updated |
|---|
No replay run yet.
Specialist research on portfolio risk, position sizing, drawdown, liquidity, stress regimes, tail risk and institutional capital discipline. Research authority does not grant trading or capital authority.
| Programme | Role | Domains | Sources | Mode |
|---|
Enter real account/portfolio state to calculate advisory risk intelligence.
Versioned rulebooks with exact source evidence. Daily loss, max drawdown, trailing/static rules, consistency, payout, leverage, news/weekend/overnight restrictions and prohibited behaviour are separated by firm and account type.
| Firm | Account | Version | Effective | Source | Rules | Change |
|---|
Mathematics, statistics, physics, chemistry, complex systems, nature, fluid/water dynamics, astronomy and experimental hypotheses are researched in isolation. Nothing reaches market decision logic until translated to a measurable feature and empirically validated.
| Candidate | Domain | Hypothesis | Gate | License | Updated |
|---|
Versioned, evidence-linked memory with critic, contradiction, temporal validity, experience and cross-domain validation. Research is not operational knowledge until Governance admits it.
Only runtime-verified agents appear in the operational view. Scheduler completion, research success, empirical validation and trading win rate are separate metrics.
| Agent | Role / Kind | Mode / Domains | Live / Replay | Status | Tasks | Orchestration OK | Validation | Latency | Last heartbeat |
|---|
Measured source assessment state. Unknown remains unknown.
| Domain | Score | Samples | Success | Failures | Updated |
|---|
Live production transport and storage telemetry.
Decision, Model, Guardian, Evidence/Data and Knowledge/Risk supervisors vote independently. A hard reject remains a veto; experience may calibrate confidence but cannot lower security, evidence or capital floors.
Continuous governed discovery across institutional, scholarly, book-catalog and public hedge-fund research sources.
| Programme | Next run | Last run | Runs | Failures | Last mission |
|---|
| Name | Provider | Type | Domain | Category | Audience | Access | Next scan | Failures | State |
|---|
Connectors remain locked until your own valid subscription, entitlements and root-only credentials are configured. QuantNova never bypasses paywalls.
Persisted system incidents, not synthetic alerts.
Storage, runtime and platform readiness.
Visible by design; unlock only after the real licensing backend is production-ready.
Cryptographic product license issuance.
—Runtime entitlement and validity verification.
—Controlled machine and installation identity binding.
—Governed activation lifecycle and revocation controls.
—Tamper-evident licensing security history.
—Protected distribution and internal logic safeguards.
—Trading performance & win rate remain unavailable until the Trading Core backend is implemented, validated and explicitly unlocked.
Locked until Trading Core produces real trade outcomes