R RayCo LabsMarket Intelligence

RayCo Research Lab

Building intelligence through evidence—not hype.

RayCo is developed through historical analysis, unseen-window testing, live market observation, independent shadow lanes, and promotion gates. This page tracks the public story of that work without exposing proprietary thresholds or pretending research output is customer-ready performance.

Current stageClosed Alpha
System modeLive Research + Shadow Validation
Core marketsBTC · ETH · SOL
Last updateJuly 2026

The standard

Research first. Promotion second. Customer exposure last.

RayCo research is designed to answer difficult questions before a signal or market-state decision reaches a customer-facing surface. Can the behavior survive realistic costs? Is it dependent on one narrow market window? Does it fail during exhaustion? Is there enough liquidity to act? Does it remain useful when evaluated on unseen data?

Important:

Research, shadow candidates, illustrative product states, and internal promotion lanes are not live trading recommendations and should not be interpreted as realized customer performance.

Public research snapshot

Real research activity. No performance theater.

These figures show the current scale of RayCo's research and validation work. They are operational research metrics—not claims of customer profitability.

Recent feature observations 996,993 Evaluated across a seven-day live Shadow V4 window.
Parallel hypotheses 10 Evaluated during the first Shadow Factory research run.
Promotion-shadow candidates 53 Unique candidate events surfaced for deeper review.
Performance publication gate 250+ Independent events targeted before headline performance claims.
RayCo's public evidence standard

Headline performance results will remain unpublished until a candidate has accumulated at least 250 independent events and passed unseen-window evaluation, realistic after-cost testing, multiple market conditions, and live shadow validation.

This is a publication standard—not a guarantee of future trading or customer performance.

What RayCo is running now

Not one experiment.
An expanding research system.

CURRENT PUBLIC MODE Read-only research + shadow validation Last verified July 12, 2026
CORE OPPORTUNITY RESEARCH LIVE SHADOW

Opportunity Survival

Evaluates whether an apparent opportunity persists long enough to remain usable instead of disappearing immediately after detection.

CROSS-MARKET CONTEXT LIVE RESEARCH

Market Context + Long-Context

Studies broader market behavior and longer-duration context beyond a single short microstructure event.

BTC-SPECIFIC PROTECTION LIVE RESEARCH

BTC Exhaustion Protection

Investigates whether strong-looking BTC movement is late, weakening, crowded, or approaching reversal.

ETH-SPECIFIC RESEARCH LIVE RESEARCH

ETH Market Research

Evaluates ETH-specific behavior rather than assuming every asset responds to identical market rules.

RESEARCH OPERATIONS AUTOMATED

Candidate Promotion Pipeline

Runs candidate generation, variant analysis, ranking, and read-only promotion shadows as a repeatable research chain.

No customer execution.

These systems observe, compare, reject, and rank research candidates. They do not place customer trades or represent promoted customer-facing performance.

Research promotion architecture

Every candidate must earn its way forward.

RayCo separates operational infrastructure, active research, evidence accumulation, and customer promotion into distinct stages.

01

Market data ingestion

OPERATIONAL

Coinbase Advanced and Binance.US market data.

02

Feature generation

OPERATIONAL

Market structure, momentum, liquidity, and cost context.

03

Parallel hypothesis testing

ACTIVE

Multiple research ideas evaluated independently.

04

Candidate ranking

ACTIVE

Promising candidates separated from exploratory output.

05

Independent shadow validation

ACCUMULATING

Read-only observation across unseen live conditions.

06

Customer intelligence

PROMOTION GATED

No candidate advances until the evidence standard is met.

PUBLIC PERFORMANCE GATE 250+ independent events

Event count alone is not enough. A candidate must also survive realistic costs, unseen evaluation windows, multiple market conditions, and live shadow validation.

Validation framework

What a candidate must survive.

01

After-cost reality

Fees, half-spread, impact, and safety requirements are treated as part of the opportunity—not an afterthought.

02

Unseen-window evaluation

Promising rules must be challenged outside the exact window where they were discovered.

03

Survival over time

RayCo evaluates whether apparent edge persists or disappears shortly after detection.

04

Market-context dependence

A candidate may only work in particular phases, alignments, or liquidity conditions.

05

Exhaustion + reversal risk

Strong-looking movement can be late, crowded, or close to failure.

06

Live shadow behavior

Read-only shadow systems observe current conditions before customer exposure is considered.

Research operations

A living record of RayCo research activity.

Independent research lanes are evaluating market context, opportunity survival, and exhaustion behavior.

July 2026

Independent shadow research expanded

Separate lanes are evaluating market context, survival, and exhaustion without modifying customer-facing output.

July 2026

Research promotion pipeline operational

Candidate generation, variant analysis, ranking, and read-only promotion shadows now run as a repeatable pipeline.

June 2026

Survival confirmation became a core research theme

Research shifted from detecting movement alone toward determining whether an opportunity remains usable long enough to matter.

Public research snapshot

Real research activity. No performance theater.

These figures show the current scale of RayCo's research and validation work. They are operational research metrics—not claims of customer profitability.

Recent feature observations 996,993 Evaluated across a seven-day live Shadow V4 window.
Parallel hypotheses 10 Evaluated during the first Shadow Factory research run.
Promotion-shadow candidates 53 Unique candidate events surfaced for deeper review.
Performance publication gate 250+ Independent events targeted before headline performance claims.
RayCo's public evidence standard

Headline performance results will remain unpublished until a candidate has accumulated at least 250 independent events and passed unseen-window evaluation, realistic after-cost testing, multiple market conditions, and live shadow validation.

This is a publication standard—not a guarantee of future trading or customer performance.

What RayCo is running now

Not one experiment.
An expanding research system.

CURRENT PUBLIC MODE Read-only research + shadow validation Last verified July 12, 2026
CORE OPPORTUNITY RESEARCH LIVE SHADOW

Opportunity Survival

Evaluates whether an apparent opportunity persists long enough to remain usable instead of disappearing immediately after detection.

CROSS-MARKET CONTEXT LIVE RESEARCH

Market Context + Long-Context

Studies broader market behavior and longer-duration context beyond a single short microstructure event.

BTC-SPECIFIC PROTECTION LIVE RESEARCH

BTC Exhaustion Protection

Investigates whether strong-looking BTC movement is late, weakening, crowded, or approaching reversal.

ETH-SPECIFIC RESEARCH LIVE RESEARCH

ETH Market Research

Evaluates ETH-specific behavior rather than assuming every asset responds to identical market rules.

RESEARCH OPERATIONS AUTOMATED

Candidate Promotion Pipeline

Runs candidate generation, variant analysis, ranking, and read-only promotion shadows as a repeatable research chain.

No customer execution.

These systems observe, compare, reject, and rank research candidates. They do not place customer trades or represent promoted customer-facing performance.

Research promotion architecture

Every candidate must earn its way forward.

RayCo separates operational infrastructure, active research, evidence accumulation, and customer promotion into distinct stages.

01

Market data ingestion

OPERATIONAL

Coinbase Advanced and Binance.US market data.

02

Feature generation

OPERATIONAL

Market structure, momentum, liquidity, and cost context.

03

Parallel hypothesis testing

ACTIVE

Multiple research ideas evaluated independently.

04

Candidate ranking

ACTIVE

Promising candidates separated from exploratory output.

05

Independent shadow validation

ACCUMULATING

Read-only observation across unseen live conditions.

06

Customer intelligence

PROMOTION GATED

No candidate advances until the evidence standard is met.

PUBLIC PERFORMANCE GATE 250+ independent events

Event count alone is not enough. A candidate must also survive realistic costs, unseen evaluation windows, multiple market conditions, and live shadow validation.

What we publish

Honest progress without giving away the engine.

Development stage

Whether RayCo is in research, shadow validation, closed alpha, or a later customer-ready stage.

Research themes

The problems being investigated, such as survival, exhaustion, market context, liquidity, and execution quality.

Aggregate learnings

What improved, what failed, and what the team learned—without publishing proprietary decision thresholds.

Promotion discipline

A clear separation between exploratory research, shadow candidates, promotion candidates, and customer-facing output.

Follow the build

We are not waiting for perfect conditions to learn.

We are building the system that learns from them.

Request Early AccessExplore Bot Intelligence