After-hours earnings shock
A public company update lands while the US primary market is closed. Tidelight starts with the question, not a prediction.
TIDELIGHTPUBLIC EVALUATOR MODE · NO LOGIN
AI TRADING DESK / AFTER-HOURS INTELLIGENCE
From an overnight event to a decision you can inspect.
Tidelight is an evidence-first control room for tokenized US equities. This fixed replay shows the complete trust path in under five minutes, without pretending that a staged result is live performance.
01 / THE FIVE-MINUTE REVIEW
Follow one decision without opening a private workspace.
Claims stay attached to sources
Publisher, timestamp, excerpt, confidence, and counter-evidence remain visible beside the conclusion.
The Bitget Reality market is mapped
The affected rToken, quote freshness, session state, and execution context are shown before a scenario is considered.
Risk gates can say no
The paper plan is only available when evidence, freshness, sizing, exposure, and pause controls agree.
02 / WHAT IS VERIFIED
Proof is part of the interface.
Every important statement has a place in the trace. A missing source, stale quote, contradictory claim, or failed gate is visible rather than silently replaced with confidence.
Open the public evidence manifest ↗01
Source lineageCaptured source → claim → counterpoint → conclusion
02
Market contextReality instrument, provider timestamp, session and cost context
03
Decision traceQuestion → evidence → exposure → scenario → risk gate
04
Safe executionPaper-only review with deterministic rejection reasons
03 / THE OTHER TWO MODES
One evidence standard across the product.
05 / REPRODUCIBLE PROOF
Inspect the artifacts, not just the pitch.
PUBLIC EVIDENCEOpen the evidence manifest.Routes, artifact status, limitations and reproduction requirements ↗LIVE RESEARCH TRACEInspect the rMSFT evidence.Owner-published source and market snapshot ↗OWNER-PUBLISHED REPLAYPublish a real research run.After a completed run, share a 30-day source-linked replay with judges ↗Publishing requires a signed-in owner and explicit consent.ALPHA FACTORYExport a reproducible strategy run.Authenticated runs include candle fingerprints, costs and holdout data ↗NIGHTWATCHExport private paper evidence.Account-scoped decisions and paper ledger; review privacy before sharing ↗
06 / OWNER-PUBLISHED RESEARCH
Seven live records. One clear evidence boundary.
Five new token reviews join the two earlier public examples. Every card opens a real owner-published snapshot with captured sources, instrument checks, and risk gates. Links expire 30 days after publication and can be revoked.
Across the five new runs, 40 issuer headlines were collected per company from 17–30 publishers. AI synthesis was unavailable in each, so these are transparent fallback records rather than completed analyst theses. No directional assessment or trade handoff was made; each replay shows quote freshness and unmeasured token-market checks.
RTSLA · EV / AUTOMOTIVEWould you review the rTSLA market?40 issuer headlines across 30 publishers. The token quote was stale; depth, spread, and premium/discount were not measured.Inspect published replay ↗RMSFT · AI / CLOUDHow do AI investment and Azure demand change the risk picture?40 issuer headlines across 23 publishers. The token quote was fresh at capture; market depth, spread, and premium/discount were not measured.Inspect published replay ↗RAAPL · DEVICES / SERVICESWhat should a long-term rAAPL review verify?40 issuer headlines across 22 publishers. The token quote was stale; the run did not assess rAAPL tradability.Inspect published replay ↗RGOOGL · REGULATION / AI SEARCHHow should antitrust remedies and AI-search competition be reviewed?40 issuer headlines across 17 publishers. The token quote was fresh at capture; liquidity, spread, and premium/discount were not measured.Inspect published replay ↗RAMZN · CLOUD / RETAILHow do AWS investment and retail economics shape the review?40 issuer headlines across 21 publishers. The token quote was fresh at capture; no trading handoff was made.Inspect published replay ↗RNVDA · TOKENIZED MARKETIs the rNVDA market worth reviewing?Earlier owner-published run. Inspect its original issuer coverage, instrument context, and stated limitations.Inspect published replay ↗NVDA + TSLA · COMPARISONCompare the latest evidence and risks.Earlier owner-published comparison with separate issuer assessments and decision gates.Inspect published replay ↗
04 / THE HONEST BOUNDARY
Useful evidence is not a promise of profit.
This route is a reproducible product evaluation. It does not claim audited alpha, investment advice, live execution, or an invented paper fill. When the real gates do not qualify, Tidelight says so.