The data behind the engine
The sharpest signal in the market sits in public — on-chain, scattered, unstructured, unused. WAYL already collects it, cleans it, structures it, labels it with real outcomes, and runs a live engine on top that proves it works. The dataset, the pipeline, and the capability are all here.
The thesis
Raw public data is cheap and chaotic. Clean, structured, continuously-updated, outcome-labeled data is rare — and it's what every model, desk and platform actually needs. WAYL was built to do exactly that: take public prediction-market, on-chain and market data and turn it into something you can build on. Not a one-off scrape — a living pipeline, running 24/7, with three months of history and growing every day.
What's in the vault
778 tracked wallets — who the sharpest capital is, what they hold, and when. On-chain, timestamped, deduped.
Every entry, exit, size and move across tracked venues — the raw behavior behind the positions.
Where sharp money agrees, captured hour by hour — so you can see conviction build and shift over time.
Signals resolved win / loss against real results. Real, labeled ground truth — the hard, expensive part for any model.
258 prediction markets, 167 crypto assets, equities, and 25K+ cross-exchange pricing edges — one pipeline, many markets.
500K+ price snapshots and 59K+ large-trade events captured via live websocket — streaming, not just historical.
Why it matters
Cleaned & structured
Deduped, typed, normalized into queryable tables — not a raw dump.
Outcome-labeled
120K+ rows tied to real results. Ground truth is the rare part.
Continuous
Three months timestamped, updating 24/7 — a living feed, not a snapshot.
Multi-domain
Prediction markets, crypto, equities — the same pipeline, extendable.
What it can be
Labeled real-world prediction and behavioral data — the clean, ground-truth inputs models are starved for.
Smart-money flow, consensus and cross-venue pricing, delivered as a structured feed your models can consume.
Where the sharp money lands before the lines move — demand, liquidity and mispricing signals for your desk.
A system that turns messy public sources into clean, documented feeds — pointed at whatever domain you need.
Beyond the dataset
The same system that built all of this can be pointed at new domains and new tasks. WAYL was built solo, with AI, entirely from public sources — which makes it fast, adaptable and cheap to extend. Need a specific dataset collected, cleaned and labeled? That's the core capability, not a side project.
Proof it's decision-grade
The clearest proof that data is clean enough to build on is that something real is built on it. WAYL's live, publicly-graded signal engine runs entirely on this data — every call timestamped before the event and settled in the open.
See the data in action — the live, interactive track record.
View the live record →Work together
Structured, documented data delivered on a schedule — historical or streaming.
A specific dataset collected, cleaned and labeled to your spec.
Think there's a fit for your models, desk or platform? Let's talk.
All data is derived from public and on-chain sources, then collected, cleaned, structured and labeled by WAYL ENGN. Row counts, time depth and breadth are drawn from the live production database. Full schemas, samples and methodology are available under NDA.
WAYL ENGN · confidential data overview · waylengn.com