Features & Benefits — Market Prism

Why this matters

It's not an LLM wrapper.

The AI-stock-tip space has filled up with thin shells around black boxes — paste a ticker, get a confident-sounding paragraph that may or may not match the actual 10-K. Market Prism is built the other direction.

The language model is one tool inside a seven-step forensic pipeline. Every narrative claim is grounded in primary-source documents — filings, earnings transcripts and company guidance — then scored across drift, VMS, decay and mass, cross-checked for coordination across outlets, and classified for exhaustion and hype-bubble traps. If the model wants to say something the source doesn't support, the pipeline catches it before it reaches your screen.

Three patents pending. The model is a tool. The forensic chain is the product.

01Multi-source ingestfilings · transcripts · guidance

02Narrative extractionclaim-level

03Source-authority scoringdomain weighted

04Multi-signal scoringdrift · VMS · decay · mass

05Coordination & spam detection60-min windows

06Narrative trap classificationexhaustion & hype bubbles

07Forensic scorecarddelivered to UI

Live screen · Daily Plays

Your market briefing. Every morning.

Conviction-ranked signals across day trades, swing setups, momentum, value picks and narrative traps — refreshed daily from the narrative scorecard.

Live data

Market Prism Daily Plays dashboard: Day Trade, Swing Trade and Momentum lanes with active picks and signal summary

/dashboard · updated each session

The full slate, tracked to settlement.

Day Trade, Swing Trade, Momentum lanes — plus an Active Today / Wins / Hit Rate summary sitting right in the header. Every play is auto-tracked through entry, exit and alpha-vs-SPY.

Conviction-ranked picks across day trade, swing, momentum, value and narrative-trap lanes
Sector intelligence rail on the right ranks where the strength is concentrated
15-minute live refresh during market hours via a four-cron pipeline

Live screen · Stock Forensics

Per-ticker, source-anchored, non-negotiable.

Open any ticker and the forensic read shows up first: hype score, fundamental disconnect, narrative bias, conviction tier, and the trade strategy implied by the setup.

Source anchored

Market Prism AAPL stock forensics page with Hype Trap badge, fundamental disconnect score and conviction rating

/ticker/:symbol

The scorecard, with the receipts attached.

The header tells you the verdict — Hype Trap, Fundamental Disconnect, Infrastructure flags — before you scroll. Below it, a 5-day move projection, an above-$250 narrative target, and a conviction percentage backed by trade rationale. Toggle the horizon to see 1D, 5D, 30D and 90D projections side by side.

Machine-learning forecast — projections trained on historical narrative-to-price outcomes, not heuristic rules
Multi-horizon view — 1-day, 5-day, 30-day and 90-day move bands rendered together
Bias / Move / Horizon / Target grid surfaces the trade thesis up top
Tabbed forensics — Trading Strategy, Company Profile, Live Market Snapshot, Fundamentals
Market Scholar sidekick is RAG-grounded on this ticker, never free-form

Audit a ticker

Live screen · The Leaderboard

Forensic edge. Every signal ranked.

15 live boards in a single screen — from Loudest Stories, Spam Alarm and Claude's Conviction List to Whale Tracks, Hidden Bargains and Today's Verdicts. Coordination alerts, manipulation risk, keyword signals and analyst accuracy refreshed daily.

15 live boards

/leaderboard

What the noise is doing — at a glance.

Where the standard dashboard shows you trades, the Leaderboard shows you the raw forensic state of the market: who's loudest, who's most over-hyped, who's quietly discounted, and where coordinated narratives are forming.

Spam Alarm flags coordinated-coverage tickers before they reach the tape
Hidden Bargains ranks the names with the cleanest fundamental setup
Narrative Traps calls out stories about to exhaust — the moment the reversal usually hits

Open the leaderboard

Live screen · Narrative Physics

The whole market, plotted on narrative mass vs. earnings clock.

Each dot is a ticker. The Y axis is how much narrative weight has accumulated. The X axis is days to its next earnings print. The Fire Zone on the right (DTE ≤ 14) is where stories meet hard catalysts.

Narrative physics

Narrative Physics chart

Where the story has nowhere left to hide.

Bubble size = narrative mass. Cyan ring = coordination ≥ 60. Dashed green ring = ideal entry. Red dots are the names where the story is breaking down faster than the price has admitted.

Fire Zone overlay isolates tickers with earnings inside two weeks
Coordination & entry rings visualize the spam-alarm and signal layers
Cluster reading — see exactly when the megacaps are crowding into the same window

Live screen · Trading Cards

Collect. Analyze. Play.

High-conviction trade signals packaged as live, ranked, scored cards. Each one is a tradeable read with a state, a command, a confidence and the underlying forensic context.

Live picks

Market Prism trading cards: Peak Narrative, Momentum Confirmed and Losing Steam cards across DASH, UNH, TSN, SNPS, INTC

/cards · live signals

Rarity = conviction. State tag = what to do now.

Filter by Unicorn, Platinum, Gold, Surprise, Bearish or Bullish. Each card carries a state — Peak Narrative, Momentum Confirmed, Losing Steam — plus a buy/hold/short command and a confidence percentage.

Sentiment-coloured top accent — green bull, red bear, cyan neutral
Live ranking across rarity tiers, refreshed throughout the session
Value Picks are tagged FYI and excluded from P&L by design

Start collecting

Live screen · Signal Lab

Price meets narrative. See the story behind the move.

Overlay narrative force, risk pressure and coordinated-campaign signals on any ticker's price — then generate an AI interpretation of what's actually driving today's action.

Narrative overlay

Market Prism Signal Lab: NVDA price overlaid with narrative force, risk pressure and meme cycle lines

/signal-lab · per ticker

Receipts, charted.

Toggle Narrative Force, Risk Pressure, Meme Cycle, Coordination overlays directly on top of the candles. The Signal Lab is where you go to verify a signal is actually adding edge — not just noise.

Multi-overlay charting — turn each forensic layer on/off independently
Analyze button generates a Market Scholar interpretation of the chart you're looking at
Multi-timeframe ranges from 1D out to YTD and MAX

Try the lab

Live screen · Narrative Lab

Validate any narrative. What's the bullshit?

Paste a headline, tweet or claim about any stock — including ones outside our tracked universe. The Narrative Lab embeds it, finds the closest historical analogues, and aggregates what actually happened next.

5.5M observations

/narrative-lab

Pattern-match against 5.5M historical observations.

Take any claim — "AMD will hit $4500 with NVDA's data center wins by 2030" — and the lab will find every analogous narrative we've seen, weight them by source authority, and tell you what actually happened to the price after.

Free-form input — works on tickers we don't actively track
Analogue aggregation beats single-shot LLM "predictions" by orders of magnitude
Source-authority weighted — a verified analyst's call counts more than a content farm's

Test a narrative

Head to head

How Market Prism stacks up against the usual stack.

Bloomberg gives you data. Yahoo gives you headlines. Seeking Alpha gives you opinions. The new crop of AI-stock apps gives you confident text. Market Prism gives you a forensic chain — with our own Source Rating, Analyst Scoring and Fair Value model. Zero sell-side, broker or PR feeds inside the system.

Capability	Market Prism	Bloomberg / Refinitiv	Yahoo Finance	Seeking Alpha	FinChat / AI stock apps	LLM wrappers & AI stock apps
Multi-signal forensic scoring	Drift · VMS · decay · mass per claim	No native scoring	No	Author opinion	Single confidence number	None
Source-anchored claims	Filings, transcripts & guidance — auto-anchored	Sources available, not auto-anchored	Headlines only	Author opinion	Sources as RAG context	Hallucination-prone
Source-authority weighting	Domain-weighted, not volume-weighted	Curated wires	Aggregated equally	Author rep system	Embeddings only	None
Coordinated-coverage detection	60-min window coordination flags	Not surfaced	No	No	No	No
Narrative exhaustion / decay model	Narrative physics scoring	No	No	No	No	No
Public paper-trading book	$1.5M, ~106 positions, tracked to alpha-vs-SPY	No	No	Author "picks" only	Some demo portfolios	No
Daily slate with hit/miss accountability	Lane-tagged, hit_date logged, public	No	No	Newsletter-level	Backtests, no daily slate	No
Forensic chain (not a generated paragraph)	Seven-step pipeline, model is one tool	Manual research workflow	No	No	RAG'd LLM output	Pure LLM output
Proprietary Source Rating	In-house domain trust score per outlet	Curated wires, no public score	None	Author rep only	Treats all sources equally	None
Independent Analyst Scoring	Our own conviction list, scored daily	Aggregates sell-side ratings	Republishes sell-side	Author opinions	Re-uses sell-side data	None
In-house Fair Value model	Built from primary sources, not licensed	BVAL / proprietary, $$$	None	Quant rating only	Imports third-party DCF	None
Free of Wall Street sell-side influence	Zero broker, banker or PR feeds	Sell-side is the product	Pulls in sell-side ratings	Sponsored content surfaces	Sell-side data dependencies	Trains on sell-side material
Patent-pending methodology	Three applications filed	Proprietary, undisclosed	No	No	No	No
Cost to retail	Free tier + retail-priced pro	$2k+/mo terminal	Free / Yahoo Plus	$240/yr Premium	$20–50/mo	$20/mo Plus

Full Partial Missing

Benefits

What you actually get out of using it.

Less time reading content-farm "analysis." More time looking at evidence. A daily slate you can audit, a forensic page per ticker, and a heatmap that shows the pressure before the move.

Edge

Front-run narrative exhaustion.

Saturation flagging tells you when a story has run out of room — historically the moment the price reverses. Most tools tell you the story exists; we tell you when it's over.

Trust

Every claim has a citation.

If the dashboard says a company is "guiding lower," you can click through to the exact line in the 8-K. No paraphrase, no GPT confabulation.

Speed

One screen, the whole forensic read.

Scorecard, source alignment, source breakdown, coordination flag, price series — all in one ticker page. No five-tab Bloomberg dance.

Accountability

Public hit/miss tracking.

$1.5M paper book, every Daily Play tagged with hit_date and alpha-vs-SPY. If our slate is bleeding, you'll see it before we do.

Filter

Domain authority over volume.

A messy post from a known industry voice outranks a thousand polished AI-generated articles. We score the source, not the word count.

Defense

Coordination flags catch campaigns.

Five outlets publishing near-identical framing inside an hour isn't consensus — it's a campaign. We surface it before you trade into it.

Real screens. Real signal. Not a chatbot.