Discover what might connect. Establish what the connection supports.
Quantum Trellis expands the field of candidate explanations. Prosdocimi determines which deserve causal standing. Many market workflows begin with variables already selected by an analyst: price, volume, earnings, rates, commodities, sector membership and sentiment. Useful relationships can remain outside that initial selection. Quantum Trellis brings qualified quantum and classical observations into a common geometric frame to surface further relationships worth investigating. Prosdocimi supplies the market context and the challenge that follows.
- Quantum Trellis: which observations deserve to be examined together?
- Tomography: which original records explain, contradict or qualify the geometric connection?
- Prosdocimi: is the relationship associational, predictively useful, intervention-sensitive or causally defensible?
- Memory: what did the investigation learn, and what would justify revisiting it?
SO WHAT? A better question can change the investigation.
A falling lumber price looks like good news for a homebuilder. Purchase commitments can change that interpretation. Follow one illustrative connection through discovery, tomography, causal challenge and memory. Each step changes what the customer can reasonably investigate next.
ILLUSTRATIVE MARKET WALKTHROUGH · Select a step to follow the same evidence.
Select a clue to bring it forward · Drag to explore · Swipe vertically to scroll
Select a sphere to inspect its evidence. A violet ring marks your selection; rose marks the contract challenge after inspection. Colour describes the view, not causal standing.
Dashed joins are hypotheses. Highlighted contract evidence challenges the initial explanation.
A commodity move becomes a question about exposure.
A lumber observation, purchasing exposure, demand conditions and a margin record appear in one candidate neighbourhood. Quantum-derived and classical geometry give the investigator a reason to examine these records together.
- Lumber observation
- What moved, in which units and over which period?
- Purchasing exposure
- Which purchases use this material?
- Demand conditions
- Could the operating regime change the interpretation?
What moves forward: A candidate graph with links to its contributing evidence.
Schematic positions and illustrative records. This interaction explains the workflow; it does not calculate an investment outcome or a causal verdict.
Prosdocimi supplies governed market evidence.
The Prosdocimi Price Index (PPI) workflow contributes classical observations from its market surfaces. Each observation retains its instrument, observation time, source, units, market regime, uncertainty, provenance, missing-data status and transformation history. These are inputs to investigation; causal standing has to be established separately.
- Sovereign Screener, Gain Access Class and Asset Class Framework.
- Gain Matrix, Prosdocimi Index, PPI Analysis and Correlation Matrix.
- Price and return behaviour, market regimes, commodity relationships and risk fields.
- Sector and industry classifications, company fundamentals and macroeconomic conditions.
- Later outcomes and user decisions, preserved with when they became available.
Every shape has a return address to its evidence.
CPU and GPU systems transform eligible classical observations into geometric representations under a qualified phase-space mapping. The original record stays available through its index. Geometry makes comparison possible while source identity, units, uncertainty and transformation choices remain explicit. A centroid alone does not contain enough information to recover an original document.
- Centroid: location in the declared representation.
- Shape and covariance: uncertainty and local structure.
- Orientation: direction under the qualified mapping.
- Trajectory: change through time.
- Neighbourhood: candidate relationships among compatible observations.
- Provenance: source records, transformation identity and the route back to both.
Quantum scouting opens questions outside the initial query.
Independently produced QPU observations enter comparison only when their geometry and the market representation share a justified frame. Repeated, qualified convergence can nominate an unexpected neighbourhood for review. The question is why these observations express compatible structure, and whether the relationship survives scrutiny. Provider diversity alone does not establish source independence; shared inputs and reconstruction methods must be accounted for.
- Unexpected market neighbourhoods and cross-asset relationships.
- Commodity-to-company pathways and hidden sector dependencies.
- Regime-sensitive structures and unusual state transitions.
- Candidate leading indicators and variables that may participate in the same mechanism.
- A new investigative question, with its QPU and classical evidence attached.
The neighbourhood becomes a candidate graph.
Trellis organises the discovery into nodes and qualified candidate edges for Prosdocimi to investigate. A lumber movement, purchasing exposure, demand regime, interest-rate change, margin transition, supplier disruption and company-price response can become distinct nodes. Geometric proximity does not establish an edge’s causal direction. Where the evidence cannot support ordering or orientation, that part of the hypothesis remains unresolved.
- Each edge carries its source evidence and temporal ordering, including uncertainty in that ordering.
- Preserve proposed direction, uncertainty, independence status and compatibility results.
- Retain transformation identity, alternative explanations and missing evidence.
- Send hypotheses for examination; keep the distinction between a proposed graph and a supported causal model.
Tomography makes the geometric lead inspectable.
Prosdocimi follows each indexed node and edge to the original records. CPU and GPU tomography examine the connected evidence from several views: time, exposure, regime, source and transformation. Here, tomography means examining a connected evidence set through declared analytical views. It is distinct from quantum-state tomography, which reconstructs a quantum state from measurement statistics. The original classical records are retrieved through their indexes, rather than inferred backwards from a displayed shape.
- Which observations and variables contributed to the convergence?
- Did the sources measure compatible quantities in compatible time windows?
- Does the relationship persist across market regimes?
- Could preprocessing, selection or a shared upstream source explain the connection?
- Which records contradict the preferred explanation?
- What additional evidence would distinguish competing mechanisms?
The Causal Ladder limits what may be concluded.
Association asks what occurs together. Intervention asks what changes when an input is changed. Counterfactual reasoning asks what would have happened under an alternative condition. The levels require different evidence and assumptions; repeating a correlation does not move it up the ladder. Prosdocimi’s investigation must declare its identification strategy and the limits of the resulting claim.
- Association: test recurrence, chance, stability across instruments and regimes, and comparison with conventional baselines. Predictive usefulness still needs evaluation on held-out future data without leakage.
- Intervention: test an identified mechanism against timing, alternative causes, placebo checks and perturbations. Sensitivity in a fitted model is not itself an observed intervention effect.
- Counterfactual: use an identified structural model, a justified comparator and explicit assumptions to estimate the alternative outcome. Preserve where that comparison becomes unreliable.
A mechanism model is a testable account of change.
Prosdocimi’s stated model is xₜ₊₁ ≈ Axₜ + Buₜ + c. Here xₜ is the declared current state, xₜ₊₁ the next state, uₜ the proposed input, A the state-transition map, B the input-response map and c an offset. Units, sampling, residual uncertainty, regime and fitted scope belong with the model. Estimating A and B from observations does not, by itself, identify a causal effect.
- Does changing the proposed input alter the next state beyond normal system evolution?
- Does the proposed cause precede the effect through a plausible transmission path?
- Do placebo tests, perturbations and competing explanations challenge the mechanism?
- Which experimental design or defensible identification assumptions support interpreting the response causally?
Every market stage adds to the same investigation.
The Prosdocimi crosswalk connects Sovereign Screener → Gain Access Class → Asset Class Framework → Gain Matrix → PPI → PPI Analysis → Correlation Matrix. Each stage records what it observed, added, contradicted and was not permitted to conclude. A later model can support, qualify or dispute an earlier finding while retaining the original record.
- Preserve evidence, derived states, candidate mechanisms and tensions.
- Keep model outputs, provenance, missing data and declared limitations.
- Record user decisions and later outcomes at the time they become available.
- Retain revisions and failed hypotheses so later reviews can reconstruct what changed.
Unsupported edges lose standing, not their history.
An edge may remain associational, intervention-sensitive, causally supported under declared assumptions, counterfactually identified, conflicted, indeterminate, insufficiently evidenced or refused. These are scoped assessments, not an automatic promotion sequence. Missing timing blocks a directional claim. Shared sources limit corroboration. A confounder can overturn the preferred explanation. An unidentified question remains unresolved. Remove an unsupported edge from the accepted causal model while preserving the proposal and the reason it failed.
The next investigation inherits evidence and lessons.
Completed investigations return their results and open questions to EcoSynQ Memory. A later investigation can prioritise a promising mechanism while retaining the right to revise it. Refused and indeterminate relationships remain useful history: the team can inspect why an attractive explanation failed and what new evidence would be needed to reconsider it.
- What Quantum Trellis surfaced and which records created its geometry.
- The proposed graph, surviving edges and rejected edges.
- The alternative explanation that prevailed and the regime in which it held.
- What Prosdocimi could conclude, what it could not conclude, and what happened later.
Better questions. Inspectable answers. A stronger next investigation.
The value is a wider field of candidates coupled to a disciplined way of testing them. Evaluate the integrated workflow against the customer’s existing analysis, including strong classical discovery methods. Measure useful new leads, false leads, review time, out-of-sample stability and total investigation cost. Those comparisons establish what the quantum contribution adds in a particular workload.
- Reduce dependence on the analyst’s initial variable selection.
- Bring commodity, company, sector, macroeconomic and operating evidence into the same qualified investigation.
- Give Causal AI focused starting hypotheses, while retaining alternative graphs and uncertainty.
- Preserve the route from every candidate to its source records.
- Learn from contradiction and refusal as well as from supported findings.
From a hidden connection to an accountable decision.
Prosdocimi supplies market evidence. Quantum Trellis represents and compares qualified observations. QPU scouting opens candidate relationships. Tomography traces and examines their records. Prosdocimi tests the candidate graph. The Causal Ladder limits the conclusion. Memory preserves what survives and what fails. This page describes the Quantum Trellis–Prosdocimi integration architecture. The interactive market walkthrough is an authored example, not a replay of a validated Prosdocimi result or a trading signal.