Phase 1 · Consumer packaging

Autonomous quant research, made investable.

A public-facing dashboard for Andrew's AI research engine: thousands of QuantConnect-tested variants, clear champion strategies, OOS-first validation, and a simple path from research transparency → paid signals → premium strategy lab.

View monetization ladderSee current champions

Updated 2026-05-17 05:17 UTC · Source files: sprint_state.json, XAU research_state.json, knowledge.db

The engine

How the product earns trust

1Generate

AI proposes one-at-a-time strategy variants from papers, prior failures, and champion lineage.

2Backtest

QuantConnect runs IS and OOS tests. Weak variants are rejected instead of hidden.

3Promote

Only candidates beating the champion under risk guardrails become new public winners.

4Package

Consumers get watchlists, model portfolios, options playbooks, and premium research feed.

Current champions

What consumers see first

Separate deployable lanes from moonshot/R&D lanes. That keeps marketing honest.

Core / deployable research

Equity L/S

Active

Long the strongest quality/momentum stocks, short the weakest, then scale risk down when market internals deteriorate.

ChampionV2301
OOS Sharpe2.038
CAGR89.7%
Max DD18.1%
Drawdown use60% of 30% cap

Edge: Quality + momentum are persistent equity factors; the research edge is the crash/dispersion exposure-control layer.

Consumer package: Best packaged as a weekly US equity long/short watchlist and regime-risk meter.

  • Iterations: 160
  • Wins / qualified candidates: 5
  • KB records: 251 · OOS-tested 120 · best OOS 2.090
QC project ↗·Backtest proof links pending

Core / income research

SPY Options Income

Active

Systematic SPY put-credit-spread engine with VIX, DTE, and drawdown discipline.

ChampionV557
OOS Sharpe2.182
CAGR65.6%
Max DD15.8%
Drawdown use63% of 25% cap

Edge: Harvests index insurance premium while strict sizing and volatility filters prevent one trade from dominating the book.

Consumer package: Best packaged as a monthly options-income playbook with exact strike/DTE/risk notes.

  • Iterations: 223
  • Wins / qualified candidates: 32
  • KB records: 149 · OOS-tested 81 · best OOS 2.215
QC project ↗·Backtest proof links pending

Diversifier / low drawdown

Global Macro ETF

Active

Cross-asset ETF rotation across equity, bonds, gold, dollar, agriculture, and cash fallback.

ChampionMA_V20
OOS Sharpe1.261
CAGR25.1%
Max DD6.1%
Drawdown use24% of 25% cap

Edge: Simple regime rotation gives consumers a less intimidating diversified sleeve versus high-vol crypto tracks.

Consumer package: Best packaged as a model-portfolio allocation dashboard.

  • Iterations: 28
  • Wins / qualified candidates: 3
  • KB records: 28 · OOS-tested 25 · best OOS 1.261
QC project ↗·Backtest proof links pending

Active alpha lab

XAUUSD Gold CFD

Active

30-minute gold CFD trend/trailing-stop research with closed-PnL validation and ongoing AlphaEvolve iteration.

ChampionXAU_V3463
OOS Sharpe3.886
CAGR140.9%
Max DD8.4%
Drawdown use24% of 35% cap

Edge: XAU_V3463 currently combines high OOS Sharpe with controlled drawdown; still research-grade until live forward test exists.

Consumer package: Best packaged as a premium gold signal/risk lab after paper-forward proof is added.

  • Iterations: 6,143
  • Wins / qualified candidates: 3,368
  • KB records: 6,143 · OOS-tested 3,755 · best OOS 3.938
  • ATR reward/risk: 2.46x flexible
QC project ↗·IS proof ↗ · OOS proof ↗

High growth / high risk

Crypto Conservative

Active · V10910 in flight

BTC/ETH trend system with volatility-regime sizing and hard drawdown controls.

ChampionV10902
OOS Sharpe5.259
CAGR688.6%
Max DD45.0%
Drawdown use100% of 45% cap

Edge: Crypto trend alpha is real, but volatility is the killer; regime-aware leverage is the differentiator.

Consumer package: Best packaged as an optional high-volatility satellite sleeve, never as the core product promise.

  • Iterations: 121
  • Wins / qualified candidates: 6
  • KB records: 281 · OOS-tested 232 · best OOS 9.965
QC project ↗·Backtest proof links pending

Moonshot / experimental

Crypto Aggressive

Active

Leverage-enhanced BTC/ETH acceleration track built to maximize upside under a very high drawdown cap.

ChampionV10872
OOS Sharpe7.780
CAGR934.4%
Max DD68.0%
Drawdown use97% of 70% cap

Edge: Useful for storytelling and alpha discovery, but too volatile for most retail consumers without strong warnings.

Consumer package: Keep behind a Pro/Advanced risk gate; market as research, not retirement allocation.

  • Iterations: 140
  • Wins / qualified candidates: 32
  • KB records: 281 · OOS-tested 232 · best OOS 9.965
QC project ↗·Backtest proof links pending

R&D / saturated

Forex R&D

Active

Daily FX breakout family has not produced a durable OOS edge yet.

ChampionFX_V11
OOS Sharpe—
CAGR11.6%
Max DD25.7%
Drawdown use51% of 50% cap

Edge: The lesson is valuable: the current family is saturated and needs a new architecture, not more parameter tuning.

Consumer package: Do not sell as a signal yet; show as transparency that the engine rejects weak markets.

  • Iterations: 78
  • Wins / qualified candidates: 0
  • KB records: 63 · OOS-tested 1 · best OOS -0.594
QC project ↗·Backtest proof links pending

R&D / rebuilding

Commodity ETF R&D

Active · CMT_V59 in flight

Commodity ETF sleeve is still below promotion gate in recent runs.

Champion—
OOS Sharpe—
CAGR—
Max DD—
Drawdown use0% of 35% cap

Edge: Potential diversifier, but latest variants are failing IS gates; keep it in lab mode.

Consumer package: Only include in the public dashboard as research pipeline proof, not as a paid signal.

  • Iterations: 55
  • Wins / qualified candidates: —
  • KB records: 53 · OOS-tested 0 · best OOS —
QC project ↗·Backtest proof links pending

R&D / early

Rates ETF R&D

Active

Rates proxy sleeve exploring duration, real-rate, and inflation regimes.

Champion—
OOS Sharpe—
CAGR—
Max DD—
Drawdown use0% of 25% cap

Edge: Useful as portfolio hedge research, but early variants remain below gate.

Consumer package: Future model-portfolio feature; not monetizable yet.

  • Iterations: 23
  • Wins / qualified candidates: 0
  • KB records: 23 · OOS-tested 0 · best OOS —
QC project ↗·Backtest proof links pending

Monetization hypothesis

Start with trust, then charge for timeliness

Audience growth

Free Radar

$0
  • Delayed champion board
  • Weekly market regime note
  • Research transparency, no trade alerts

Retail power users

Pro Signals

$49–99/mo
  • Current model watchlists
  • SPY options-income playbook
  • Gold/crypto risk meter
  • Telegram alerts

Serious traders / small funds

Premium Lab

$299–999/mo
  • Live research feed
  • Full backtest evidence
  • Strategy packs
  • Priority feature votes

Advisors / communities

White Label

$5k+ setup
  • Branded dashboard
  • Custom universe
  • Private alerts
  • Monthly research call

Live research tape

Recent engine activity

  • 2026-05-17T05:15:17Z | XAU_V6139 no promotion. Best valid OOS return remains 140.892% (Sharpe 3.886, DD 8.4)
  • 2026-05-17T05:16:38Z | XAU_V6140 OOS: WR=42.0 Sharpe=0.742 DD=1.8 Orders=257.0
  • 2026-05-17T05:16:38Z | XAU_V6140 no promotion. Best valid OOS return remains 140.892% (Sharpe 3.886, DD 8.4)
  • 2026-05-17T05:16:47Z | XAU_V6141 OOS: WR=50.0 Sharpe=3.595 DD=8.7 Orders=258.0
  • 2026-05-17T05:16:47Z | XAU_V6141 no promotion. Best valid OOS return remains 140.892% (Sharpe 3.886, DD 8.4)
  • 2026-05-17T04:49:47Z | crypto_aggressive | V108884 | IS=1.641/70.584%/81.4% bt=104962d5cf491f7007c3574f34b72d5d | OOS=0.900/35.804%/53.2% bt=73d2d7dd16d2f8a813909135db828387 | miss_below_champ
  • 2026-05-17T05:03:40Z | forex | FX_V72 | is_fail 0.000<0.3
  • 2026-05-17T05:05:31Z | macro | MA_V28 | IS_PASS(0.565) -> OOS
  • 2026-05-17T05:05:34Z | macro | MA_V28 | OOS_RUNNING | bt=534f027d0ece96f648a46328902c65e6
  • 2026-05-17T05:07:15Z | macro | MA_V28 | IS=0.565/8.594%/15.9% bt=0759d5bf588e092a0e5123ea0b24b125 | OOS=0.943/20.664%/6.1% bt=534f027d0ece96f648a46328902c65e6 | miss_below_champ
  • 2026-05-17T05:08:59Z | options | V606 | is_fail 0.481<1.0
  • 2026-05-17T05:10:45Z | rates | RT_V23 | is_fail -0.004<0.4

Risk posture

What we do not oversell

!Backtest ≠ live PnL

Phase 1 sells research transparency. Phase 2 needs paper-forward/live-forward proof before aggressive claims.

!Crypto is not core

High CAGR tracks are visually powerful but must sit behind risk gates and drawdown warnings.

!OOS is king

Avoid promoting IS-only wins. Show misses to prove the engine rejects bad ideas.

!No financial advice

Dashboard is educational/research tooling unless licensed advisory/compliance is built.

Educational and informational research dashboard only. It does not constitute financial, investment, tax, or trading advice. Backtests are hypothetical and can overfit; live execution, slippage, borrow, fees, liquidity, taxes, and psychology can materially change results.