AlphaHound icon
AlphaHound · Cross-dataset market intelligence

What changed.
What it means.
Whether to act.

A daily briefing for the stocks you watch — built by crossing institutional positions, insider trades, dark pool volume, options flow, congressional disclosures, and macro data into signals no single source can give you.

Free Fully local No accounts No subscription macOS · Windows · Linux
🖥 Detecting your platform…
Download AlphaHound Detecting platform…
~30 MB · no account required · data downloads in the background after first launch

Opens every morning to this

The Briefing is a ranked feed of what actually moved the needle for your watchlist — ordered by severity, each card with a plain-English "why it matters" and a one-click path to dig deeper or ask the AI.

The Briefing
Wednesday 7 May · 7 signals
All (7) Your watchlist (4) High Medium Insider Dark pool Options
High NVDA
3 new institutional positions opened this quarter while dark pool volume ran 2.4× above the 30-day baseline
Why it matters · New fund entries alongside unusual off-exchange volume suggests coordinated accumulation ahead of a catalyst. Options chain confirms elevated call activity in the same window.
Open ticker Ask AI → Source
Med MSFT
CFO made a $4.2M open-market purchase — first insider buy in 18 months
Why it matters · Executives rarely buy at market unless they believe the stock is undervalued relative to near-term expectations. Cross-checks with a net increase in institutional positions this quarter.
Open ticker Ask AI → Source
Info AAPL
2 congressional disclosures filed this week — purchases totalling $15K–$50K reported
Why it matters · Weaker signal in isolation, but congress trades have historically preceded sector-level moves. No corroborating institutional change this quarter.
Open ticker Ask AI → Source

One level deeper, one click away

Every Briefing card links to a full view. Here's what's waiting.

Smart Money · Win Rate ▾
1Citadel Advisors
91%
2Millennium Mgmt
79%
3Two Sigma
73%
4Renaissance Tech
69%
5Bridgewater
62%
Smart Money Leaderboard
Which funds have the best win rate when they open new positions? Ranked by historical accuracy across every 13F quarter in your database.
Conviction Scores · Watchlist
NVDA
88
MSFT
74
META
67
AAPL
41
TSLA
22
Conviction Scores
A 0–100 composite from 5 independent data sources per ticker: institutional flows, dark pool spikes, insider activity, momentum, and accumulation persistence.
Financials Healthcare Energy Consumer Technology Industrials Utilities Q4 → Q1 sector rotation
Money Flow · Sankey
Visualises institutional capital rotation between sectors across quarters. Band thickness = confidence. Scrub history to watch money move over time.
Dark Pool Volume · 52 weeks · NVDA
Low → Spike
Dark Pool · Spike Calendar
52-week heatmap of off-exchange volume z-scores per ticker. Spot the weeks where dark pool activity ran unusually hot — often weeks before price moves.
Upcoming · Your watchlist
13F filing deadline — Q1 2026 May 15
NVDA insider cluster · 6 purchases Apr 28
MSFT earnings disclosure window Apr 30
META material event · 8-K filed May 2
Filing Calendar
Upcoming 13F windows, insider purchase clusters, earnings dates, and material events — all scoped to your watchlist tickers.
Why is dark pool activity in NVDA elevated this week?
Based on your local data: 3 new institutional positions opened in Q1 alongside a 2.4× dark pool spike. This pattern — new 13F entries concurrent with elevated ATS volume — historically precedes a price move in 68% of cases in your database…
AI Chat · Local LLM
Ask questions about your signals in plain English. The AI has full context — your watchlist, the filings, the dark pool data. Runs on your machine via Ollama. Nothing leaves.

From install to first briefing in under 5 minutes

No account creation. No API keys for the free path. No configuration required to get started.

1

Download & install

Native app for your OS. Double-click, drag to Applications (macOS) or run the installer. No dependencies.

2

Pick your watchlist

Search for tickers. AlphaHound backfills the last two years of filings and data for each one in the background.

3

Open The Briefing

Every morning, a ranked feed tells you what changed overnight — what it means, and whether to act. That's the whole loop.

What makes a signal here different from Yahoo Finance

Every card in The Briefing joins at least two independent datasets. An insider purchase on its own is noise — but an insider purchase that coincides with new institutional positions and elevated dark pool volume is a pattern. That cross-dataset join is what AlphaHound does. No signal surfaces from a single source.

Institutional positions × insider trades Dark pool spikes × 13F entries Options flow × congressional disclosures Macro shifts × institutional exits

All publicly available. All free.

No subscriptions, no data vendor fees. Everything AlphaHound uses is legally available to anyone — it just takes a machine to join it all together.

Institutional holdings
Quarterly filings from all funds managing ≥$100M — who holds what, how much changed
Insider trades
Regulatory disclosures filed within 2 days of every executive buy or sell
Dark pool volume
Weekly off-exchange trading volume per venue — the part of the market not visible on public order books
Options flow
Front-month call/put ratios versus 30-day baseline — unusual positioning ahead of moves
Congressional disclosures
Mandatory trade filings by US representatives and senators — historically well-timed
Material events
Regulatory filings for leadership changes, investigations, M&A, and other market-moving events
Macro data
Interest rates, yield curve, inflation, M2 — used to contextualise institutional positioning changes
ETF holdings
ARK, iShares, Vanguard snapshot positions — cross-referenced with institutional 13F entries

System requirements

Native desktop app built with Rust and WebKit — not Electron. Starts fast, stays fast.

🍎
macOS
10.15 Catalina or later
Intel · Apple Silicon
Universal binary
🪟
Windows
Windows 10 or 11
64-bit only
.msi installer
🐧
Linux
Ubuntu 20.04+ · Debian 11+
.AppImage · .deb
x86_64
4 GB RAM minimum
8 GB recommended · 16 GB if running larger local AI models
~2 GB disk (initial sync)
Grows with history depth · local AI model adds 2–10 GB (optional)

Local AI via Ollama is optional — the Briefing and all signals work without it. AI adds the conversational "ask why" layer on top.

Copied to clipboard