100K HF Book
$100,000
Active research
The flagship hedge-fund-posture research account. Same brain, same gates, same discipline you would use with size — built to be watched in public.
Free access until June 15
An autonomous trading desk in your pocket.
AstroTrades is a private intelligence room: SJ live in WhatsApp, the AstroTrades Brain managing research books in public, an institutional-grade options-flow stack, and a futures bot lab — all engineered to surface real edge, not noise.
Educational research, analytics, and community. Not financial advice. Trading futures and options involves substantial risk.
AstroTrades Brain
Live command feed
GREEN
Multi-book
10 daily slots
Strict gating
Latest posture
Discipline first. Pass reasons are part of the product.
The brain runs ten times a day, every weekday. It either takes the trade with a clear thesis, watches with a defined trigger, or passes — and tells the room why.
Watching
Continuation longs on confirmed flow
Trigger requires hold above structural level + matching call sweeps
Watching
Failed-rally fades into known walls
Invalidation: reclaim of resistance with positive flow turn
Risk gate
No forced trades
When conviction or event-risk does not clear, the brain passes — out loud.
Three live research books
Four-account model bake-off
Millions of flow rows ingested
Continuous backtesting + retros
Watch the work
Three research books. One futures bot lab. All in public.
Members get a live look at the AstroTrades Brain managing three sandbox accounts and the futures bot running across four model variants. Every entry, every pass, every reason — visible.
10K Aggressive Book
$10,000
Active research
Higher conviction floor, tighter R:R, faster ladder. For setups where size matters more than signal volume.
1K Challenge
$1,000
Protected ladder
A small-account scaling story. Only A+ alignment, only when the math works. No forced trades just to make noise.
Futures Bot Lab
Multi-account
Live + paper
Four independent runtimes, same engine, different decision models — running side-by-side so the best judgment wins.
The systems
Sophisticated enough to matter. Simple enough to act on.
Every member message is the output of a stack: scanners, gating, model decisions, and a human-readable translation. You see the call. You see the why. You see what would change the read.
01
AstroTrades Brain
Our AI-driven HF-style decision engine — model-agnostic by design — that converts market state into actions, watchlists, pass reasons, and risk posture across three sandbox accounts.
02
Risk Reversal Scanner
Finds financed institutional options structures, then validates direction, balance, catalyst risk, and collar fingerprints before anything reaches the room.
03
Flow Hunter
Ranks unusual options activity by flow score, timeframe, opening confidence, strike band, and OI confirmation — not just by dollar size.
04
SJ in WhatsApp
A live group-chat analyst that explains market reads, summarizes setups, and translates machine output into something you can act on.
05
Intraday Narrator
Turns live tape, macro releases, gamma exposure, and breaking news into useful commentary instead of generic recap filler.
06
Futures Bot
Trigger-driven futures execution with hard stops, dynamic sizing, and a qualitative supervisor — AI in advisory, never in the hot loop.
How we find edge
The work happens before the message ever lands.
Edge is engineered, not guessed. We pull from institutional-grade data partners, score in parallel across multiple frontier AI labs, walk every strategy forward against years of history, and gate every alert through a structured fact-check before it reaches the room.
01
Ingest at scale
Millions of flow rows, options chains, prints, dark-pool aggregates, macro releases, and news items pulled and normalized every session.
02
Score in parallel
Multiple decision models from leading AI labs run the same problem in parallel. We compare them, score them, and let the best judgment win — every week.
03
Backtest before we believe
Every new strategy is walked forward against hundreds of thousands of historical setups before it ever gets near a live book.
04
Gate before we publish
Every alert passes a structured fact-check, conviction floor, and style gate before it reaches the room. No spam. No hopeful prints.
Risk Reversal Scanner
The banger engine — without exposing the playbook.
Most flow tools blast every sweep. Ours waits for same-expiry financed structures with the right OTM placement, notional balance, opening evidence, catalyst context, and collar-defense fingerprints — then validates each one before SJ ever writes a message.
Detect
Find paired institutional structures with enough balance and size to deserve attention.
Validate
Reject hedging collars, duplicates, incoherent regimes, weak catalysts, and bad structure math.
Translate
Turn the surviving structure into strikes, thesis, risk map, and what would weaken the read.
Distribute
WhatsApp first. Cross-posted to public channels only when the call still passes the style gate.
The stack
Built on what the desks use. Tuned for what traders need.
We partner with the data and AI providers institutions actually trust — multi-vendor by design, so we’re never one outage or one model away from going dark. We don’t name our specific vendors publicly; that’s part of the moat.
Data spine
- Tier-1 options-flow telemetry
- Institutional-grade market data
- Real-time gamma & dealer positioning
- Macro, earnings & filings feed
Decision layer
- Frontier AI labs (multi-provider)
- Live model bake-off across four accounts
- Continuous A/B against millions of historical setups
- Domain-trained classifiers for flow & news
Execution + risk
- Hard stop enforcement
- Per-book posture engine
- Circuit breakers & sector caps
- Broker-truth P&L reconciliation
The room
A WhatsApp group that actually feels like a desk.
SJ is in the room with you. Ask about a setup, a level, a catalyst — get an answer that makes sense. The point isn’t to follow blind alerts; it’s to learn how the system thinks so you can think the same way.
- Morning brief before the open with the day’s map
- Real-time alerts when scanners trigger high-conviction setups
- SJ in the room to explain WHY a setup matters, in plain English
- Every alert ships with trigger, invalidation, R:R, and what would change the read
- Position updates and posture shifts posted as they happen
- Weekly retros: what worked, what missed, what the system is learning
Founding-member window
Free until June 15. Then $150/mo.
Founding members get the room while we harden the system in public. After June 15, membership funds the data feeds, the model usage, the monitoring, and the continued buildout — same room, same SJ, same edge.