Find the feature. Understand the result. Know what to do next.

Make the daily desk yours

Your daily desk opens after sign-in. It combines the current published story window with your own reading preferences and saved context.

• Personalize: choose Executive for significance and next steps, Analyst for facts and evidence assessment, or Media for reporting pace and coverage. Select topics and add comma-separated keywords to influence relevance ordering. Other topics remain visible.
• Read and save: mark stories as read or unread. Save a story to keep its evidence snapshot after it leaves the published window, and add a private note. Removing a saved story also removes that snapshot and note.
• Search and filter: search titles, places and source names; filter by topic or source; sort by relevance, latest evidence or source breadth. Topics without stories stay disabled.
• Verify the evidence: open original source links or full analysis. Source counts describe coverage and are not a guarantee of independent corroboration.
• Export brief: download up to 12 stories from your current filtered view as Markdown, with available source links and your private notes. Review the contents before sharing.

Start Here

The 60-second QRMonitor workflow

• Sign in and choose the question or workspace that matches what you want to know.
• Set country and language scope only when narrowing the evidence is useful.
• Read the Takeaway or Assessment before technical detail.
• Check What Changed or movement since the previous cycle.
• Open Sources or Evidence for statements that matter most.
• If the issue matters later, create a story watch, alert or Synthesize investigation.
• QRMonitor can send email alerts according to your watch while you are away from the platform.

What each main area is for

• Analyses: ready-made monitoring questions across major domains.
• Synthesize: structured investigation of a person, group, technology, object, event, geography, time period or concept. Availability can be controlled by your administrator.
• Watchboard: saved stories and changes after you started watching them.
• History: a Then vs Now record of saved stories, choices and observed outcomes.
• Alerts: email notifications for meaningful evidence transitions.
• BYO / Custom monitoring: persistent monitoring for questions outside the ready-made catalog.

Sign In and Navigate

Signing in

• Open qrmonitor.ai/auth/#signin and use your registered email and password.
• If you forget your password, use the OTP/password-recovery flow.
• Your private watches, Synthesize investigations, BYO monitors and history stay associated with your account.

Top ribbon

• Analyses: return to the post-login analysis home.
• Synthesize: open structured investigations when enabled for your account.
• Watchboard: saved stories, active watches, material changes and next steps.
• History: compare a saved story with its latest matching state.
• Alerts: create and manage evidence-change notifications.
• Profile: account information and password/account controls.
• Knowledge Center: open this guide at any time without needing to leave the product permanently.

Use Ready-Made Analyses

Choose the analysis that most closely matches the decision or situation you want to understand.

Country and language scope

• Use country filters when geography is part of the question. A narrow filter can hide relevant external reporting.
• Selected languages affect retrieval, not only display or translation. Evidence may appear in its original selected language.
• Several languages can reveal different reporting frames and strengthen local-source coverage.

When to use Ask Your Own Question

• Use a ready-made Analysis for a broad monitoring domain.
• Use BYO for a focused query you want to monitor persistently over time.
• Use Synthesize for assessment, hypotheses, indicators, claims, relations and open questions.

Read the Analysis Cockpit

Read in this order

• Takeaway: shortest statement of what the current evidence means.
• Recommended posture: suggested response level based on evidence and decision synthesis. Treat it as guidance, not instruction.
• Confidence: how strongly available evidence supports the current assessment. Confidence is not severity.
• Needs attention: priority developments meeting the cockpit's urgency/attention logic.
• Independent sources: breadth of separate reporting contributing to the analysis.
• Activity pulse: how active the live signal is; high activity does not by itself prove real-world severity.
• Signal constellation: where pressure or activity is concentrated across analysis modules.
• Action compass: why the development matters, what to do next and what to watch.
• Evidence stream: developments currently driving the picture, with direct source access.
• Movement since last cycle: direction and size of change from the previous cycle.

Source ecosystem

• Agreement: how much evidence converges on similar claims or direction. High agreement can still reflect common sourcing.
• Disagreement / contradiction: materially different or conflicting claims; inspect evidence rather than automatically rejecting the story.
• Largest source share: how concentrated the source picture is in the biggest contributor.
• Coverage footprint: where reporting comes from and which languages are represented.

Go deeper

• Evidence Lab: trace sources, module findings and provenance.
• Decision Radar: deeper decision-oriented movement, convergence, priorities and signal change.

Watch Stories, Use Watchboard and History

Watch a story

• Open the story that matters and choose Watch story.
• The story title and saved assessment become the watch context; the parent monitor remains the stream QRMonitor checks for later reporting.

Story watch conditions

• Any important change
• Escalates / gains urgency
• De-escalates / cools
• Diminishes / drops from coverage
• Receives an official update
• Gains independent support
• Becomes disputed
• Returns after going quiet

Change labels

• Escalating: stronger urgency or worsening direction.
• De-escalating: cooling or becoming less urgent.
• Diminishing: reporting or activity is dropping away.
• Expanding: broadening in evidence, geography or coverage.
• Updated: material change without a simple escalation/de-escalation label.
• Stable: no material change against the saved state.

Watchboard and Story History

• Watchboard: saved stories, changed stories, active watches and suggested next steps.
• Then vs Now: compare the saved story with the latest matching story.
• You can record simple choices and outcomes to create a practical decision trail.

Create Alerts and Notifications

Quick, advanced and guided alerts

• Quick alert: use a common preset.
• Advanced setup: choose a metric, comparator and trigger when you understand them.
• Guided alert: ask for meaningful change, stronger evidence, urgency, disagreement, an official source or renewed activity without choosing a numerical threshold.

Query Watch

• Appears or disappears from coverage.
• Escalates, de-escalates or changes materially.
• Receives an official update.
• Gains independent support or becomes disputed.
• Returns after going quiet.

Topic Watch

• New, hot/accelerating, fading or removed.
• Disputed or confirmed by stronger evidence.
• Spreading across sources/regions, officially updated or returning after quiet.

How matching works

Query and Topic watches use contextual relevance, so semantically related reporting can match even when wording differs. After relevance is selected, transitions are evaluated from current-versus-previous evidence.

Create a Synthesize Investigation

Use Synthesize when you need a defensible structured investigation rather than a normal monitoring summary. Synthesize may be enabled only for selected accounts.

Choose a subject type

• Person: role, actions, affiliations, counterparties, exposure and constraints.
• Group / Organization: leadership, operations, financing, partnerships and dependencies.
• Technology: adoption, suppliers, bottlenecks, regulation, substitutes and economics.
• Object / Asset: ownership, location, condition, capacity, availability and dependencies.
• Event: actors, timeline, triggers, response, escalation and consequences.
• Geography: economy, policy, trade, energy, infrastructure and external dependencies.
• Time / Period: sequence, before/after change, lead/lag and temporal clustering.
• Concept / Issue: actors, mechanisms, competing interpretations, policy/economic effects and trends.

Optional Context

Optional Context is the analytical question or scenario you want Synthesize to answer. Make it specific. For a conditional scenario, the system should treat the condition as hypothetical unless evidence shows it actually occurred.

Analytical lenses

• Connections: identifies people, organizations, places, events, technologies or concepts linked to the subject and highlights strong evidence-backed connections.
• Relations: explains the nature of connections, such as ownership, cooperation, conflict, dependency, association or operational interaction.
• Causal hypotheses: tests possible explanations while separating evidence from unproven causal assumptions.
• Change and trajectory: shows what is changing over time and whether the subject appears to strengthen, weaken, escalate, stabilize or shift.
• Influence and dependency: identifies what can affect the subject, what it depends on, and possible vulnerabilities or leverage points.
• Narrative divergence: detects materially different or conflicting descriptions across sources, languages, regions or groups.
• Geographic diffusion: shows how an event, narrative, risk or development spreads, distinguishing story geography from reporting geography.
• Early warning: looks for emerging signals, unusual changes or evidence combinations that may indicate a developing risk, disruption or escalation.

Synthesize Assessment, Confidence and Z-score

Assessment

• Judgment: main conclusion Synthesize can defend from the current evidence and analytical objective.
• Because: most important evidence-backed reasons supporting the judgment.
• Falsifiers / This assessment breaks if: observations that would materially weaken or overturn the current assessment.

Confidence

Confidence is normally shown as Low, Moderate or High, sometimes with an ICD-style probability word. It describes the state of the evidence, not the importance of the subject.

Confidence drivers

• Source volume: whether enough relevant evidence is available.
• Source independence: how much evidence comes from genuinely separate source families.
• Corroboration depth: how many independent families support promoted claims.
• Language spread: whether selected-language evidence is broad or concentrated.
• Geographic spread: whether evidence covers relevant geography broadly enough.
• Run stability: whether the substantive finding remains stable across saved versions.
• Engine agreement: convergence between deterministic and provider-assisted evidence-backed findings.

Coverage Z-score / momentum

• Definition: how far today's context-relevant coverage is above or below its recent daily baseline.
• z = -0.6 means coverage is 0.6 standard deviations below the recent average; it does not mean 60% lower.
• A high coverage Z-score means unusual reporting activity. It does not by itself prove the underlying condition is more severe.
• Baseline: the recent comparison window used to understand what is normal for the same subject/context.

Investigation Progress and Resolve Gap

Investigation Progress

A workflow-completeness indicator showing how far the investigation has progressed. It is not a truth score and is not the same as confidence.

States

• Initial: first evidence-backed picture.
• Developing: useful evidence exists but important gaps remain.
• Substantiated: evidence, independence and hypothesis separation are stronger.
• Watchable: enough structure and watch conditions exist for continued monitoring.

Progress factors

• Evidence coverage
• Independent corroboration
• Hypothesis separation
• Open-question closure
• Watch readiness

Open Questions and Resolve Gap

• Open Questions are ranked collection gaps that could materially improve the investigation.
• Resolve Gap launches a targeted evidence pass while preserving the original analytical objective.
• A click does not automatically mark the question resolved; the next evidence-backed version must actually close it.

Assessment Changes

• Initial: first version.
• Stable: substantive direction unchanged.
• Changed: main judgment materially changed.
• Strengthened: direction stayed similar while confidence/evidence improved.
• Weakened: direction stayed similar while confidence/evidence deteriorated.

ACH and Challenge Assessment

What ACH means

ACH = Analysis of Competing Hypotheses. It compares several plausible explanations against the same diagnostic evidence instead of asking only what supports a preferred explanation.

ACH cell meanings

• ++ strongly consistent.
• + consistent.
• · neutral / not diagnostic.
• - inconsistent.
• -- strongly inconsistent.
• ? evidence not yet collected or not yet scoreable.

Diagnosticity and weighted inconsistency

• Diagnosticity: how useful an evidence item is at separating competing hypotheses.
• Weighted inconsistency: credibility-weighted evidence that conflicts with a hypothesis. Lower means the hypothesis is less refuted by current evidence.
• The top-ranked hypothesis is the least-refuted explanation in the current evidence scope; it is not automatically true.

Challenge Assessment

• Identifies evidence that weakens the current leading hypothesis.
• Lets you inspect an alternative and evidence that favors it over the leader.
• Unknown ACH cells can become targeted collection opportunities.
• Challenge mode is read-only: only new evidence can change the production assessment/ranking.

Indicators, Warnings and Evidence Discovery

Indicator Watchboard

Indicators convert warnings into named, stateful conditions that can persist across Synthesize versions.

• Tripwire: checkable condition that moves an indicator when new evidence meets the rule.
• Armed: watched but not fired.
• Tripped: tripwire condition met.
• Cooling: fired earlier but recent evidence shows less movement.
• Reset: returned to a non-triggered state.

Early warning

A case-specific warning about evidence conditions that may deserve attention. It is evidence-bounded and not a deterministic prediction of the future.

Evidence Discovery moments

• Assessment direction changed.
• A tripped indicator appeared.
• An open question was resolved.
• Independent corroboration increased.
• A new relation-map entity appeared.
• A hypothesis changed rank.
• High-diagnostic evidence was identified.
• Contradictory or contested evidence appeared.

Optional Analyst Challenge

If enabled, you can review high-diagnostic evidence before seeing hypothesis ranks, select the least-inconsistent explanation, record confidence/rationale, then compare with deterministic ACH ranking.

Relation Map and Exploration

The Relation Map is a stable 2D view of the subject, observed entities, inferred analytical constructs and hypotheses. Fixed lanes help keep epistemic states separate between runs.

Epistemic status

• Observed: directly grounded in retrieved evidence.
• Inferred: analytical construct derived from evidence but not stated as one direct fact.
• Hypothesis: proposed explanation or causal possibility, not an observation.

Node and edge

• Node: a person, group, place, asset, subject, analytical construct or hypothesis.
• Edge / relation: relationship between two nodes. Open edge evidence to see why QRMonitor connected them.

Exploration actions

• Investigate this: open direct evidence that placed the node on the map.
• Why connected? inspect evidence from relationships directly connected to the node.
• Follow a relation: move through neighboring entities and relations without losing your place.
• Pivot Synthesis: start a new investigation from a selected node/relation while carrying the original scope/objective.
• Exploration trail: breadcrumb of entities/relations followed during exploration.

Claims, Diffusion and Collection Gaps

Claim Ledger

• Claim: atomic statement extracted from evidence and compared across sources.
• Reports: number of retrieved documents carrying the claim.
• Independent source families: genuinely separate reporting families after near-duplicates/syndication are grouped.
• Echo ratio / amplification: how much report volume is repetition rather than independent corroboration. High amplification does not mean false.

Claim states

• Corroborated: supported by several independent families without material contradiction.
• Single-source: rests on one source family.
• Amplified: widely reported but much volume traces to few originating families.
• Contested: materially conflicting evidence exists.
• Numeric conflict: incompatible figures for the same measure.

Diffusion

• The sequence in which a story first appears, enters new source families, languages or geographies and changes frame.
• Displayed time is normally first observed in retrieval/publication time, not necessarily real-world event time.
• Story geography: where the story is materially about.
• Source geography: where the publisher/source is located.

Collection gaps, silence and health

• Collection gap: important evidence the analysis could not obtain, often ranked by how much it could improve hypothesis separation.
• Silence: expected coverage that is absent or drops unexpectedly; consider retrieval errors, time zones and publication patterns.
• Collection health: retrieval completeness, such as errors, languages attempted/returned and evidence limits.

BYO, Account Features and Optional Tools

BYO / Custom Monitoring

• Use BYO when the ready-made catalog does not match your exact persistent monitoring need.
• Define a custom category/subject and one or more focused question streams.
• QRMonitor uses those questions for semantic retrieval and analysis, then presents the familiar cockpit/evidence experience.
• Keep questions narrow enough that new evidence has a clear reason to match.

BYO vs Synthesize

• BYO: persistent custom monitoring over time.
• Synthesize: structured investigation producing a judgment, competing hypotheses, indicators, claims, open questions and evidence exploration.

Profile, privacy and export

• Profile contains account information and authenticated password controls.
• Private results and user-specific state are isolated to the signed-in account.
• Where Export is available, use it for briefing/offline review and note the result time/version because later refreshes may materially change the assessment.

Glossary

ACH

Analysis of Competing Hypotheses; tests several explanations against the same evidence.

Assessment

The main evidence-bounded judgment for the current scope.

Baseline

Recent comparison period used to define what is normal for a metric.

Confidence

Strength of support in available evidence, not importance or severity.

Corroboration

Independent support for the same material claim.

Diagnosticity

How strongly one evidence item separates competing hypotheses.

Diffusion

Observed sequence in which a story spreads across sources, languages or geographies.

Echo / amplification

Many reports that trace back to a small number of originating source families.

Epistemic status

Whether something is observed, inferred or hypothesized.

Evidence family

A group of reports treated as one originating/near-duplicate source family for independence analysis.

Falsifier

An observation that would materially weaken or overturn the current assessment.

Hypothesis

A proposed explanation, not an established fact.

Independent source

A source adding genuinely separate reporting rather than repeating another item.

Momentum

How current reporting activity compares with a recent baseline.

Open Question

A collection gap whose resolution could improve the investigation.

Story cluster

Semantically similar articles treated as the same underlying storyline for analysis.

Tripwire

A checkable condition that moves an indicator when evidence meets it.

Weighted inconsistency

ACH score of credibility-weighted evidence that conflicts with a hypothesis; lower means less refuted.

Z-score

Number of standard deviations a value is above or below its baseline mean.