Research workflows
Replay tagged market windows instead of rebuilding them.
Research workflows need repeatable packet sets: state, tags, reasons,
confidence, evidence, replay, rows, export receipts, and clear claim
boundaries. Etna Alpha gives the notebook or review system a packet to
compare, not just another market-data pull to normalize.
Research without Etna
The first week goes into interpretation plumbing.
The team still has to reconstruct windows, define labels, attach
evidence, build replay handles, and keep export jobs tied to the
same event context.
collect -> normalize -> label -> explain -> export
Research with Etna
Tagged windows are already packaged for review.
Research can compare packet behavior across symbols, windows, and
workflows while staying inside read-only market-state context.
packet set -> replay -> export -> compare
Research object
packet set
- Window
- Fragile depth watch with replay context.
- Labels
- Depth, pressure, crowding, and evidence posture.
- Rows
- Normalized companion rows can be joined beside the packet.
- Exports
- Research jobs remain tied to account-scoped receipts.
- Boundary
- Context and review only, not trading advice.
{
"research_job": "packet_window_export",
"symbol": "ETHUSD-PERP",
"state": "fragile_depth_watch",
"packet_set": {
"window": "reviewable",
"replay": "ready",
"rows": "joinable",
"exports": "account_scoped"
},
"claim_boundary": "market-state context only",
"plan_fit": "public_research"
}
Best first loop
Research is for replay, exports, and packet-set comparison.
Builder proves the object. Pro runs active workflows. Research is
the first serious plan when longer replay, exports, and active-universe
launch coverage are the point of the purchase.