A research decision is not an executed trade.
Your trading desk
What BLUM is doing, what it could trade next, and what its paper results actually support.
Is the engine working?
Persisted worker observations; a running process is separate from an executed trade.
Worker activity has not been established by the available response. Trading results below are separate evidence.
Heartbeat and job timestamps describe worker activity. They do not certify quote freshness or a profitable trading edge. Visible views refresh automatically after one minute.
Checking the research record
Loading the actual stored evidence. Availability and missing data stay visible.
Inspect research callsUnavailable
Source-reported positions in this market.
Legacy history; qualification is evaluated separately.
Only aligned, qualified observations enter this sample.
The current registry records research decisions. It does not establish a copyable lead time; do not treat historical entries as upcoming opportunities.
What is stopping a qualified result?
The most frequent exclusions in the prospective record.
Missing evidence is not counted as a successful result.
A decision may have more than one exclusion.
A validated edge needs mature, cost-aware outcomes and a relevant benchmark. Activity and model explanations alone do not establish it.
Reading research state…
The workspace will show the actual stored objective when it is available.
A cycle count measures activity, not improved trading performance.
The paper result, with the right context
Qualified monetary results are grouped by account currency.
This is unavailable evidence, not a zero return. The Trades view retains the separate source history and its cost details.
Descriptive source history. It is not a verified prospective track record.
Inspect trades and costs