The product

Four engines, one reconciled call, and the argument behind it.

AI extracts structure from language. Deterministic math produces every number. That split is why a score can be re-derived tomorrow and come out the same — and why we can show you the arithmetic instead of asking for trust.

News Impact

10 min market hours · 15 min off-hours

Reads every article that crosses our news feeds for a ticker, ranks them, and feeds the top 50 into the aggregate score.

ReadsThe article body — up to the first 6,000 characters — not just the headline, weighted by source tier.
MethodAI extracts catalyst type, scope and evidence strength — fixed formulas produce the number.
DecayImpact fades on a set schedule, so a week-old headline can't hold a score up.
SpilloverA TSMC article also scores NVDA, routed through a sector affinity matrix.
LimitsPaywalled sources unread, opinion and reprints down-weighted.

Technical Analysis

daily, after 4:00 PM ET

Fully quantitative, zero AI. Price structure computed rather than interpreted, then adjusted for the prevailing market regime.

ReadsFull OHLCV history across daily and weekly timeframes.
ComputesTrend, RSI, support and resistance, Bollinger state, volume confirmation, catalyst readiness.
RegimeBull, bear, sideways or crash shifts the thresholds — a breakout in a crash tape isn't a breakout in a bull tape.
GuaranteeDeterministic: the same candles give the same score today and next year.
LimitsNeeds sufficient history. Currently mid-recalibration.

Crowd Focus

every 3 hours

Attention and sentiment on X, measured rather than modeled. It answers one question: is this move already crowded?

ReadsPublic posts mentioning the ticker — volume, velocity, positive/negative split.
ReportsAttention level from normal to extreme, plus sentiment tilt. Counts, not model output.
WeightExactly zero. Context beside the call, never inside it.
LimitsX only for now. Not a contrarian indicator and not presented as one.
FlagAn extreme reading on an already-crowded name is a risk flag, never an entry.

Fusion

several times per trading day

Reconciles all three under one fixed rule set and writes the argument in plain English.

RuleTechnicals set base direction. News is a bounded modifier — shades, never flips — and deliberately asymmetric.
Evidence supportFrom layer agreement, data availability and freshness. Weak agreement returns Limited Support, not a rounded-up Strong.
WritesDirection, the named setup, evidence support, drivers, risk note, and the level that breaks the read.
AuditEvery output logged, timestamped and versioned to the engine build that made it.
LimitsStale layers lower evidence support — the gap is stated, not filled with confidence.

The output

Every read carries the same six things.

Not a score in isolation. A short written assessment you can argue with — which is also what makes it checkable a week later.

01

Direction

Five core states from Strong Bullish to Strong Bearish, plus explicit at-risk flags. Never a price target, never a recommendation.

02

The setup, named

Trend continuation, Trend reversal risk, Momentum extended, Awaiting setup. The engine names the structure it thinks the stock is in — not just which way it leans.

03

Evidence support

How much data actually stands behind the call — layer availability, freshness, cross-layer alignment. Not a probability.

04

Each engine's own score

News, Technical and Crowd shown separately, so you can see where the call came from — and where it didn't.

05

Written synthesis

Plain English on how the conflict was resolved and which layer led. This is the part no chart gives you.

06

Risk note & levels

What would break this read, and the specific prices at which it breaks. Stated before the fact, not after.

Inside the app

Four screens. That's the whole surface.

01 · Dashboard

All 150+ tickers, one row each

Every row carries the Fusion call, its risk level, the named setup behind it and the three engine sub-scores — with the market's temperature and average score sitting above them. Sort by the Fusion column and the names where the engines disagree surface first; that is the whole morning routine.

Filter by sector, direction or freshness. Watchlist pins the names you hold.

Omnisc dashboard

02 · Ticker detail

The full read, layer by layer

Each engine's own score with its reasoning, then the Fusion synthesis that reconciles them: which layer led, why the others didn't, what would break the call, and the specific level at which it breaks.

Expand any engine to see the inputs it used and when they last refreshed.

Ticker detail

03 · News

The news ledger, per ticker

All news for one company in one place — every article that crossed the feeds for the ticker, not just the ones that made the score.

Cross-ticker influence is explicit: when another company's news moves this name — a supplier's print, a customer's guidance — the spillover is shown and attributed rather than silently blended in.

News screen

04 · Intelligence Feed

News with a score attached, not a headline wall

Every article that entered a ticker's top 50 appears here with its impact score, catalyst type and which tickers it touched — including the ones it reached through sector spillover.

Filter to one ticker to see exactly what the News engine read today, and what it discarded.

Intelligence feed

Coverage

150+ tickers, chosen because news actually moves them.

A curated universe beats a crawled one. These are names where headlines, contracts and supply shocks produce measurable price response — which is the only condition under which a news engine earns its place.

Coverage expands toward 200+ in 2027. Requests from subscribers are weighted first.

AI & Technology

Semis, hyperscalers, data-centre supply chain, enterprise software.

NVDA · AMD · ASML · PLTR · AAPL

Defense

Primes and suppliers where contract awards and budget cycles move price.

RTX · LMT · GD · NOC · LHX

Materials

Copper, gold, lithium and rare earths — supply headlines with direct read-through.

FCX · NEM · MP · ALB · SCCO

Spillover routing

A sector affinity matrix means a TSMC article also scores NVDA. Coverage is wider than the ticker list suggests.

On accuracy — the honest version

Every assessment Omnisc has produced is logged, timestamped, versioned against the engine build that made it, and tracked through forward windows against what the market actually did.

We are not publishing a hit-rate number yet. The Technical engine is mid-recalibration, and a figure measured against a build we're replacing would be marketing, not evidence. When calibration closes, the number appears here with its sample size and forward window next to it.

Refresh schedule you can check

News Impact10 min market hours · 15 min off-hours
Technical Analysisdaily, after 4:00 PM ET
Crowd Focusevery 3 hours
Fusionseveral times per trading day

Every card carries a freshness badge, so you always know which of these last ran.

Read the engines on your own tickers for a week.

No charge during the trial · cancel in one click · informational only, not financial advice

The product

Four engines, one reconciled call, and the argument behind it.

AI extracts structure from language. Deterministic math produces every number. That split is why a score can be re-derived tomorrow and come out the same — and why we can show you the arithmetic instead of asking for trust.

News Impact

10 min market hours · 15 min off-hours

Reads every article that crosses our news feeds for a ticker, ranks them, and feeds the top 50 into the aggregate score.

ReadsThe article body — up to the first 6,000 characters — not just the headline, weighted by source tier.
MethodAI extracts catalyst type, scope and evidence strength — fixed formulas produce the number.
DecayImpact fades on a set schedule, so a week-old headline can't hold a score up.
SpilloverA TSMC article also scores NVDA, routed through a sector affinity matrix.
LimitsPaywalled sources unread, opinion and reprints down-weighted.

Technical Analysis

daily, after 4:00 PM ET

Fully quantitative, zero AI. Price structure computed rather than interpreted, then adjusted for the prevailing market regime.

ReadsFull OHLCV history across daily and weekly timeframes.
ComputesTrend, RSI, support and resistance, Bollinger state, volume confirmation, catalyst readiness.
RegimeBull, bear, sideways or crash shifts the thresholds — a breakout in a crash tape isn't a breakout in a bull tape.
GuaranteeDeterministic: the same candles give the same score today and next year.
LimitsNeeds sufficient history. Currently mid-recalibration.

Crowd Focus

every 3 hours

Attention and sentiment on X, measured rather than modeled. It answers one question: is this move already crowded?

ReadsPublic posts mentioning the ticker — volume, velocity, positive/negative split.
ReportsAttention level from normal to extreme, plus sentiment tilt. Counts, not model output.
WeightExactly zero. Context beside the call, never inside it.
LimitsX only for now. Not a contrarian indicator and not presented as one.
FlagAn extreme reading on an already-crowded name is a risk flag, never an entry.

Fusion

several times per trading day

Reconciles all three under one fixed rule set and writes the argument in plain English.

RuleTechnicals set base direction. News is a bounded modifier — shades, never flips — and deliberately asymmetric.
Evidence supportFrom layer agreement, data availability and freshness. Weak agreement returns Limited Support, not a rounded-up Strong.
WritesDirection, the named setup, evidence support, drivers, risk note, and the level that breaks the read.
AuditEvery output logged, timestamped and versioned to the engine build that made it.
LimitsStale layers lower evidence support — the gap is stated, not filled with confidence.

The output

Every read carries the same six things.

Not a score in isolation. A short written assessment you can argue with — which is also what makes it checkable a week later.

01

Direction

Five core states from Strong Bullish to Strong Bearish, plus explicit at-risk flags. Never a price target, never a recommendation.

02

The setup, named

Trend continuation, Trend reversal risk, Momentum extended, Awaiting setup. The engine names the structure it thinks the stock is in — not just which way it leans.

03

Evidence support

How much data actually stands behind the call — layer availability, freshness, cross-layer alignment. Not a probability.

04

Each engine's own score

News, Technical and Crowd shown separately, so you can see where the call came from — and where it didn't.

05

Written synthesis

Plain English on how the conflict was resolved and which layer led. This is the part no chart gives you.

06

Risk note & levels

What would break this read, and the specific prices at which it breaks. Stated before the fact, not after.

Inside the app

Four screens. That's the whole surface.

01 · Dashboard

All 150+ tickers, one row each

Every row carries the Fusion call, its risk level, the named setup behind it and the three engine sub-scores — with the market's temperature and average score sitting above them. Sort by the Fusion column and the names where the engines disagree surface first; that is the whole morning routine.

Filter by sector, direction or freshness. Watchlist pins the names you hold.

Omnisc dashboard

02 · Ticker detail

The full read, layer by layer

Each engine's own score with its reasoning, then the Fusion synthesis that reconciles them: which layer led, why the others didn't, what would break the call, and the specific level at which it breaks.

Expand any engine to see the inputs it used and when they last refreshed.

Ticker detail

03 · News

The news ledger, per ticker

All news for one company in one place — every article that crossed the feeds for the ticker, not just the ones that made the score.

Cross-ticker influence is explicit: when another company's news moves this name, the spillover is shown and attributed rather than silently blended in.

News screen

04 · Intelligence Feed

News with a score attached, not a headline wall

Every article that entered a ticker's top 50 appears here with its impact score, catalyst type and which tickers it touched — including the ones it reached through sector spillover.

Filter to one ticker to see exactly what the News engine read today, and what it discarded.

Intelligence feed

Coverage

150+ tickers, chosen because news actually moves them.

A curated universe beats a crawled one. These are names where headlines, contracts and supply shocks produce measurable price response — the only condition under which a news engine earns its place.

AI & Technology

Semis, hyperscalers, data-centre supply chain, enterprise software.

NVDA · AMD · ASML · PLTR · AAPL

Defense

Primes and suppliers where contract awards and budget cycles move price.

RTX · LMT · GD · NOC · LHX

Materials

Copper, gold, lithium and rare earths — supply headlines with direct read-through.

FCX · NEM · MP · ALB · SCCO

Spillover routing

A sector affinity matrix means a TSMC article also scores NVDA. Coverage is wider than the ticker list suggests.

Coverage expands toward 200+ in 2027. Requests from subscribers are weighted first.

On accuracy — the honest version

Every assessment Omnisc has produced is logged, timestamped, versioned against the engine build that made it, and tracked through forward windows against what the market actually did.

We are not publishing a hit-rate number yet. The Technical engine is mid-recalibration, and a figure measured against a build we're replacing would be marketing, not evidence. When calibration closes, the number appears here with its sample size and forward window next to it.

Refresh schedule you can check

News Impact10 min market hours · 15 min off-hours
Technical Analysisdaily, after 4:00 PM ET
Crowd Focusevery 3 hours
Fusionseveral times per trading day

Every card carries a freshness badge, so you always know which of these last ran.

Read the engines on your own tickers for a week.

Start 7-day free trialSee pricing

No charge during the trial · cancel in one click

Informational only · not financial advice