A research desk for traders who check the evidence.
Find setups and catalysts, read the market with every figure dated, and keep what your pod learns. Source-linked AI and repeatable studies are being built in the shared research environment.
| Ticker | Setup | Price | Chg | RVOL | RS | 30 days |
|---|---|---|---|---|---|---|
| SMP1Sample semiconductor | Breakout | 84.20 | +3.41% | 2.3 | 98 | |
| SMP2Sample software | Breakout | 41.86 | +5.12% | 3.1 | 97 | |
| SMP3Sample security | Breakout | 301.55 | +2.18% | 1.8 | 94 | |
| SMP4Sample industrial | Breakout | 112.34 | +1.95% | 1.6 | 91 | |
| SMP5Sample networking | Breakout | 98.70 | +2.64% | 1.7 | 93 |
- 5 scanner families
- Breakout, episodic pivot, parabolic, relentless, Stockbee-style
- 8 calendar layers
- Candidates, alerts, earnings, dividends, IPOs, supply, macro, news
- Research AI in progress
- News recaps cite articles; the research chat is being built
- Private by default
- Your lists and notes stay yours; findings are shared on purpose
One desk for finding, checking and remembering.
Tools use different datasets and update schedules. Compare their sources and as-of times: a filing date, a price date and a live overlay describe different observations.
- Catalysts by dayOne calendar for graded candidates, earnings, IPOs, supply events, macro releases, alerts and news, with a session-prep list for the next open. This display filter does not change the scope of session prep, briefs or AI.Missing events are marked unconfirmed, not empty.
- Studies you can re-runIn progressThe research platform is being built around versioned definitions, dated inputs and reports that distinguish checks run from checks not run. Shared study storage and repeatable member workflows are still in progress.Historical studies, not live results.
From a question to a finding the pod can repeat.
A finding is only useful if someone else can run it again. Each step keeps what the next one needs: the scope of the question, the versioned definition, the snapshot, the report. This is the workflow the research platform is being built around; parts of it are still in progress.
- 1QuestionAsk about a ticker, a day or a setup, scoped to dates and datasets.
- 2DefinitionTurn it into rules: universe, entry, exit, costs. Saved with a version.
- 3Run on a snapshotThe run is pinned to a dated data snapshot, so it can be repeated.
- 4ReportResults, the checks that ran, and the raw data, downloadable.
- 5FindingShare it with the pod, discuss it, or archive it with its evidence.
Read a study before you believe it.
The planned research report brings together its sample, cost model and the checks actually run. The example below shows the intended presentation; it is not a completed study or a live result.
- Commissions, spread, slippage and borrow, when the study's cost model includes them
- Exits examined on winners and losers separately
- Downloadable report and data packages for supported study workflows
- Cost modelRan
- Random-entry baselineRan
- Day-block bootstrapRan
- Split-half stabilityRan
- Capacity by size tierNot run
- Multiple-testing adjustmentNot run
Private by default. Shared on purpose.
Planned research storage will keep saved questions and draft studies private, with explicit sharing for published definitions. That workflow is still in progress.
Only you
- Watchlists, portfolios and private notes
- Alert rules and where they are delivered
Your pod
- Morning briefs and the team feed
- Ideas you post and the findings you share, with their evidence
- Trade reviews of what the desk prepared and engaged with
From raw data to the next session.
Coverage differs by dataset; each page states its range and as-of time. When something is stale, partial or missing, the panel says so rather than showing a blank or a zero.
- 01CollectPrices, fundamentals, filings and calendars are loaded into our own warehouse.
- 02CheckDaily checks record stale or missing inputs, and the affected panels say so.
- 03Scan and studyScans run after the close. Member study workflows are still being completed.
- 04DeliverShort list, briefs and alerts for the next session, each figure with its source and as-of time.
What SwarMind does not do
- It does not place trades or connect to a broker.
- It does not give investment advice or recommendations.
- Studies are historical; they are not live or guaranteed results.
- AI output is research context, not a price prediction.
Members sign in here.
Access is by invitation. Your first page is the dashboard: the regime, what changed since the last session, and your short list.