The System, Fully Explained
A structured introduction to B4_UR_IZ® / Virtual Time Machine: index, spiral taxonomy, 2D–3D linkage, ring structures, time dimensionality, drill-down, hot-synced tables, taxonomic cross-reference, audit integrity, ROSE modes, projection, analytics layers, and multi-domain application.
The Virtual Time Machine (VTM Studio – Integrated) is not merely a visual layer placed over data. It is an indexed analytical environment that converts structured data into observable architecture. Instead of forcing the user to read rows, filter reports, and mentally reconstruct meaning, the system renders the data as a spatial, cross-referential, time-aware model that can be examined directly.
Its core proposition is simple but powerful: hierarchy becomes containment, magnitude becomes geometry, lineage remains visible, and time becomes a controllable dimension. The result is a shift away from interpretive labor and toward immediate observational power. A user does not need to infer where concentration, imbalance, dependency, distortion, or emergence lies; the system makes those conditions observable.
This applies to financial structures, but it is not confined to them. The same logic extends to education systems, housing-price structures, organizational hierarchies, survey frameworks, operational networks, and any dataset that possesses formal structure across periods. The engine is structure-first. The domain changes; the observational method does not.
Traditional tools flatten reality. They present multi-layered systems as rows, columns, tabs, and filters. VTM does the opposite. It preserves the natural architecture of a dataset and renders it in spatial form. Parent–child relationships are not implied by indentation or hidden inside nested spreadsheets. They are visible as a coherent structure.
Each node becomes a geometric element within a larger whole. The observer can see how top-level systems break into categories, subcategories, accounts, metrics, or other components. This allows the human brain to do what it is naturally good at doing: perceive shape, asymmetry, density, concentration, change, and relationship. The system therefore uses human visual intelligence as a primary analytical instrument, rather than reducing analysis to repetitive numerical inspection.
The index is one of the deepest strengths of the system. It is not a convenience field and not merely a naming scheme. It is the governing structure that gives every node a persistent identity, lineage, location, aggregation role, and drill path. Without the index, the model would be visually interesting. With it, the model becomes coherent, navigable, auditable, and cross-referential.
Each indexed node can carry multiple simultaneous meanings: its location within the hierarchy, its parent statement or category, its metadata, its time-series values, and, in financial applications, its mapped regulatory or taxonomic reference. Because the same indexed identity drives the 2D map, the 3D structures, the tables, the tooltips, the drill logic, and the exports, every expression of the system stays synchronized.
This is why the system can maintain one-to-one correspondence between what the viewer sees and the underlying data source. The user is never looking at an isolated picture detached from the data. They are looking at an indexed architecture whose components can be corroborated, correlated, and traced throughout the system.
The 2D spiral is the structural map of the system. The root sits at the center, and each successive depth expands outward across radial space. Angular allocation reflects subtree weight and branching behavior, so the map does not become arbitrary decoration. It becomes a navigational instrument that preserves hierarchical truth.
The reason the structure tends toward a spiral is mathematical, not stylistic. Real hierarchies expand unevenly across depth. When these branching ratios are distributed through polar coordinates, the geometry naturally produces logarithmic spiral behavior. That is why the spiral feels organic: it is the structural consequence of the data’s own topology.
The spiral matters because it lets the observer see the whole tree, locate density, identify imbalances, build spatial memory of the model, and jump directly to the node or branch of interest. It is the cartographic layer of the platform. The 3D model shows embodied behavior; the spiral shows where everything lives.
The 2D index and the 3D model are not separate experiences. They are two synchronized expressions of the same indexed system. Selecting a node in the 2D spiral can highlight, isolate, or drill the corresponding element in the 3D environment. Selecting a node in 3D can locate it in the spiral and in the associated tables. This is a critical scaling feature because very large trees remain understandable when the user always has both a map and an embodied model.
This cross-illumination is one of the practical strengths of the system. It allows a viewer to move between overview and detail without losing orientation. The system does not force the user to “start over” cognitively each time they change view. Understanding accumulates instead of resetting.
The ring model is the primary embodied form of the hierarchy. The innermost ring represents the highest-order categories or statements. Each outward ring shows the children of the previous level. Nodes can be rendered as arcs, petals, columns, or alternative structural geometries while still remaining tied to the same index and lineage logic.
In financial usage, the center may begin with Balance Sheet, Income Statement, Cash Flow, Equity, KPI, or other primary structures. In education, it may begin with district, grade, subject, section, or assessment branch. In housing or macroeconomic models, it may begin with country, state, metro area, index, or component class. The engine does not require one particular domain; it requires structured hierarchy.
The point of the 3D form is not novelty. It is that structure, scale, and relation become visible at once. A user can see how branches widen or compress, how magnitude propagates, and where dependency clusters concentrate.
Time is one of the defining distinctions of the system. In VTM, time is not treated as a passive label, static axis, or afterthought column. It is physically embedded into the model as a stack of layers or frames. Each period occupies its own level, allowing the observer to see the system in and over time.
This means the viewer can perceive continuity, expansion, contraction, inversion, volatility, and persistence directly through the structure. Instead of comparing disconnected period snapshots, the user observes the historical form of the system as a temporal architecture.
Equally important, the time window can be controlled. The user can trim the visible range, isolate a specific segment, loop inside a selected window, or scrub through the history. This allows a viewer to focus only on the period that matters, whether that is a crisis interval, an acceleration phase, a policy regime, a housing cycle, a school performance shift, or a corporate restructuring period.
The system is not limited to past and present observation. Each node can be evaluated through its own historical behavior profile and calibrated forward. This means projection can be developed at the node level, not merely imposed from the top down. Parent structures can then update as their children evolve, allowing a bottom-up emergent forward view of the architecture.
This matters because not every branch behaves the same way. Some nodes exhibit stable growth, some mean-revert, some decay after spikes, some recover after compression, and some track ratio-bounded behavior. The system can therefore apply different modeling logic to different node types based on their own structural and temporal characteristics. The future is not guessed uniformly; it is extended from the historical identity of the indexed node.
The various ROSE structures provide different analytical perspectives on the same indexed data. They are not ornamental skins. They alter how angular or spatial allocation is handled so that the viewer can ask different questions of the same system.
ROSE Angle (Equal) allocates equal angular space to siblings, which is useful for like-for-like comparisons and clean peer scanning. ROSE Subtree allocates width according to descendant weight, which reveals structural complexity and influence. ROSE-C shifts the model into a radial skyline or footprint-like field of columns, emphasizing current concentration of value, with height encoding magnitude.
Because all of these remain tied to the same data identity, the user can move among multiple structural expressions without breaking continuity. This is one of the system’s strengths: one dataset, many coherent observational lenses.
Drill-down in VTM is not mere zoom. It is indexed re-centering. When the user drills into a branch, the model is reconstructed around that node as a new focal root while preserving lineage, temporal continuity, and cross-reference connection back to the larger system.
This allows the user to isolate an entire Balance Sheet, the whole Income Statement, a specific DuPont branch, an individual state inside a national housing model, a district inside a K–12 structure, or a consolidated wealth branch inside a multi-entity view. Large trees become manageable because the user can focus on exactly the section that matters without losing the surrounding structural truth.
The system’s tables are not secondary reports bolted on beside the model. They are an essential corroborative layer. They remain synchronized to the underlying indexed data and to the visual structures through hot-sync linkage. A selected node can highlight its corresponding table row; a selected row can focus the associated visual element. This one-to-one correspondence gives the observer confidence that what is being seen is exactly what is being represented in the source data.
These tables support exploration, audit, sorting, filtering, inspection, and verification. They make it possible to move fluidly from high-level visual understanding to precise data-level confirmation. That is one of the reasons the system is strong: it does not ask the user to trust the picture. It allows the picture to be corroborated continuously by the data layer beneath it.
Another major strength is that the system remains in sync through time and structure. Roll-ups, cross-costing relationships, computed schedules, and supporting tables are not disconnected fragments. The platform maintains structural continuity so that a figure can be traced across hierarchy, across tables, and across time. In financial use, this means audited or auditable figures can be maintained in a way that preserves integrity instead of obscuring it.
The importance here is practical. Users can test, inspect, corroborate, and reconcile. That makes the system more than a conceptual visualization environment. It is a serious analytical instrument that can preserve confidence while still enabling faster cognition.
For financial data, the index can tie directly into SEC / XBRL taxonomy. This means a node is not only positioned in the local hierarchy; it can also remain connected to a formal reporting concept. The result is that the model preserves both visual intuition and semantic rigor.
This is strategically important. It allows companies, statements, schedules, and note structures to be normalized into a common indexed framework while still retaining reference back to the regulatory taxonomy. Comparison improves. Drill-down remains conceptually grounded. The system becomes both observational and referential.
On top of the structural engine sit deeper analytical functions: DuPont decomposition, valuation logic, Monte Carlo simulation, scenario layers, ratio analysis, bookmarks, tooltips, observatory modules, and AI-assisted pattern recognition. These are not separate applications. They are layers that operate inside the same indexed environment.
That means the user can move from structural observation to forward analysis, from concentrated value inspection to valuation reasoning, from branch-level anomaly detection to projected scenario review, all without leaving the model. The platform does not fragment the analytical process; it integrates it.
The system has already shown applicability beyond traditional financial reporting. It can render house-price structures so that one state can be viewed in relation to the entire United States. It can render K–12 educational structures so that districts, grades, subjects, and performance categories can be observed in and over time. It can be extended to operational systems, surveys, organizational mapping, and any indexed hierarchy that evolves through periods.
This flexibility is critical because it proves the concept is not tied to one niche use case. It is a general method for turning structured data into observable form.
The manuals matter because they demonstrate maturity. The system is supported not only by conceptual explanation but by operational guidance: controls, workflows, module explanations, view logic, procedures, and usage pathways. A serious platform needs both a system manual and an operational manual, and this project has been moving toward that level of completeness.
That documentation strengthens the proposition for analysts, investors, collaborators, and technical reviewers because it shows the platform is not being presented as an abstract idea. It is being presented as an increasingly disciplined and documented analytical environment.
The deepest claim of the system is that it compresses cognition. It reduces the time between data and understanding. Instead of reading tables, cross-checking charts, and reconstructing relationships mentally, the observer sees structure, time, concentration, and dependency directly.
That is why the phrase observational excellence fits. The system combines human visual intelligence with indexed logic, synchronization, and AI-assisted support to create a mode of analysis that is more immediate, more coherent, and more structurally grounded than the conventional dashboard-and-spreadsheet stack.
B4_UR_IZ and the Virtual Time Machine do not merely display data. They make structured reality directly observable.
How to Begin Using the System
This section is designed to remove hesitation. It shows a new user how to begin with the embedded Morgan Stanley model, how to load additional data later, how to use mouse and touch to establish physical orientation, how to switch from the default rows into structure, how to drill into major branches, and how to move among the system’s three principal panels and statistics views.
There are two immediate ways to begin. The first is to use the embedded Morgan Stanley model already loaded within the current system. The second is to load another structured dataset of your own after joining at the membership level that gives you access to downloadable files and additional working models. A new user should start with the embedded model first, because everything is already in place and ready for movement, inspection, and drill.
- Start instantly with the embedded Morgan Stanley model: a fully built financial hierarchy already resident in the system.
- Load your own data next: anything from a simple balance sheet to a multi-tiered operating structure, home price index, valuation model, or other hierarchical time-based dataset.
- Move before you analyze: the user should first establish tactile confidence with the structure and then proceed to statistics, inspection, and deeper analytical functions.
The embedded Morgan Stanley model is not merely a demonstration. It is an active working structure with approximately 325 indexed nodes across 19 annual periods from 2006 through 2024. It allows a first-time observer to test the very capabilities described elsewhere in this document: hierarchy, time, drill-down, 2D index navigation, ROSE views, tabular correlation, statistical overlays, and structural interpretation.
Why begin here? Because the model is already calibrated. The hierarchy is complete. The time dimension is visible. The branch structure is rich enough to show the true strength of the system, yet familiar enough that users can orient themselves around major financial statements such as Balance Sheet, Income Statement, and Cash Flow.
A new viewer can therefore interact immediately without first worrying about formatting, imports, or data preparation.
Within the embedded model, the user can observe the full architecture and then hone into one branch at a time: an entire Balance Sheet, an entire Income Statement, a specific valuation component, a net-worth-oriented branch, or a note schedule. The system is already loaded with enough complexity to let the user feel how structure, time, and drill operate together.
Once the user has become comfortable with the embedded model, the next step is to load another dataset. The system supports simple and complex hierarchies alike. A user may import a single statement, a multi-tiered financial model, a KPI tree, a home price index, an educational performance structure, or another indexed dataset that unfolds through time.
| Type of Dataset | Typical Example | What Happens in the System |
|---|---|---|
| Simple Statement | Single Balance Sheet or Income Statement | Immediate rendering into structured rows, index paths, and 2D/3D views. |
| Multi-tier Financial Data | BS / IS / CF / Equity / KPI hierarchy | Full structural model with drill, time layers, roll-ups, and statistics. |
| Non-financial Indexed Data | House Price Index, K–12 results, survey structures | The same engine applies: hierarchy becomes structure, magnitude becomes geometry, and time becomes visible. |
The core message to the user is simple: the system is not limited to one kind of information. If the data has a structure and a time dimension, the system can render and interrogate it.
The first real step is physical. The user must make contact with the model. Once that occurs, the system ceases to feel abstract. The hand begins to understand the structure.
- Click + drag to rotate the 3D structure.
- Mouse wheel to zoom toward or away from the model.
- Right-click + drag or the platform’s secondary navigation gesture to pan and reposition.
- Switch the system to Mobile Mode when using touch control.
- Use one or two-finger gestures, depending on platform settings, to rotate and reposition.
- Pinch to zoom and adjust scale.
The goal here is not speed. It is familiarity. A new user should move the model slowly, rotate it, zoom into it, and then pull back from it. That first tactile connection is the takeoff point.
Once movement begins, orientation follows. The user should look for the large organizing branches first: top-level statements, major categories, or other dominant branches depending on the dataset. The purpose is to understand where the model begins and how it unfolds.
- Recognize the outer and inner structure.
- Observe that time is layered and can be trimmed or scrubbed.
- Notice that the structure is not a chart but a navigable architecture.
- Become aware that the index and branch logic remain consistent as the viewpoint changes.
At this point, the user begins to stop "reading" the data and starts to "inhabit" it spatially.
In the current version, the system may open into a rows-oriented or table-oriented representation. That is acceptable as a starting point, but the user should be encouraged to move quickly into the 2D index and the 3D structural views, because that is where the system’s full observational power becomes obvious.
Rows show the data. Structure shows the meaning.
The system allows the user to move from the current row-oriented representation into the 2D spiral index or directly into 3D. Once there, the user can compare the tabular representation with the embodied structure and begin to correlate one with the other.
The 2D index acts as the map. The 3D structure acts as the inhabited form. Both are tied by the same indexed hierarchy. A user should switch between them early so that the underlying correspondence becomes obvious.
After orientation comes drill. The first drill action should be simple and familiar: select an entire Balance Sheet, Income Statement, valuation branch, net-worth branch, or another major component. From there, descend into the structure.
- Select a top-level branch to isolate it and re-center the structure around it.
- Continue downward to major categories, subcategories, and leaf-level details.
- Watch how the system preserves lineage, time, and index position during drill.
This is the moment where the user experiences the difference between merely zooming into a picture and structurally reconstituting a model around a new point of inquiry.
The system is not one panel pretending to do everything. It is a coordinated environment of principal working areas, each with its own role and increasing analytical depth.
| Panel | Function | What the User Should Do First |
|---|---|---|
| Primary Panel | Main structure and active interaction area: movement, time, selection, drill, observation. | Rotate, zoom, select a major branch, and establish contact with the model. |
| Second / Statistical Panel | Deeper numerical interpretation, calculated analytics, and supporting metrics. | Switch here after initial exploration to confirm visually observed patterns with numerical evidence. |
| Supporting / Detail Panel | Rows, index references, tables, inspectors, and complementary views. | Use it to correlate specific nodes, statements, and values one-to-one with the structure. |
A new user should not be told merely that multiple panels exist. They should be encouraged to use them in sequence: first contact, then confirmation, then depth.
The right-hand side provides additional view and statistics access. Depending on the active version and panel, the user can open statistical information directly from the right-hand controls or move into the secondary panel for a deeper statistical layer. This matters because the system is not only visual. It is continuously tied back to the underlying figures.
When a node is selected, its corresponding numerical context can be surfaced through tables, inspectors, or statistics panels. In this way, visual observation and numerical confirmation remain in sync. The user sees the structure and simultaneously corroborates what is being seen through the figures themselves.
For a first session, the following sequence is recommended:
- Open the embedded Morgan Stanley model.
- Rotate the structure and zoom in and out until spatial orientation is comfortable.
- Switch from rows into the 2D index and then into the 3D structure.
- Select a major statement branch such as the Balance Sheet or Income Statement.
- Drill into that branch and observe how the system re-centers while preserving lineage and time.
- Open the statistical or right-side support views to corroborate what is being observed.
- Return to the broader structure and then test another branch or another render mode.
Most analytical systems begin by confronting the user with commands, menus, and complexity. This system should be introduced differently. The user should be encouraged to begin by moving, selecting, and drilling. That is how understanding begins here.
The embedded model proves that the system is already working. The import pathway proves that it is extensible. The panels prove that it is layered. The statistics prove that it is grounded. And the tactile interaction proves that the structure is not merely being viewed — it is being entered.
How the B4_UR_IZ® Engine Operates
This section shows the operating chain of confidence inside the system: indexed structure, blockchain-like parameterization, continuous audit verification, full drill-down and drill-back through the index, time continuity at every level, and forward projection that propagates through the hierarchy instead of sitting outside it.
The system does not begin with charts. It begins with structured input. Every node is assigned an indexed identity, a parent, a depth, a role, and a time series. That indexed lineage tells the engine exactly where each element belongs before anything is rendered.
This matters because the data is not treated as a pile of independent rows. It is treated as a formal architecture. Once the structure is loaded, the engine knows how every node relates to every other node and how every roll-up must behave.
The system is blockchain-structured in principle in the sense that every element is bound to its defined position, its lineage, and its parameters. A node is not free-floating. It is context-bound. Its meaning depends on where it sits in the indexed chain and how it connects to parent and child relationships.
That parameterization gives the model discipline. Each node carries identity, metadata, hierarchy, time behavior, and relationship rules. Because those rules are explicit, the engine can maintain structural continuity and prevent the model from becoming a loose visual abstraction.
This is one of the most important confidence features in the platform. The system performs a form of internal audit continuously. Values are not merely displayed. They are tested through their relationships. Parent totals, child contributions, cost flows, and cross-costing interactions can be reconciled through the structure itself.
In practical terms, the user is not being asked to trust a picture. The model is proving itself as it operates. The audit function is available throughout the system, so at every stage the observer can corroborate that the structure holds, that the costs roll properly, and that the relationships remain intact.
Once the indexed structure is established, the engine converts it into spatial form. Magnitude becomes scale. Depth becomes ring position or spatial nesting. Relationships become visible adjacency, containment, or footprint. The same underlying structure can then be rendered through rings, ROSE variants, radial fields, blocks, or other spatial modes.
The purpose is not decoration. The purpose is to make structural truth visible. Concentration, imbalance, dependency, cost pressure, and comparative scale become things the eye can apprehend immediately.
The 2D index, the 3D model, the table view, and the audit logic all point to the same indexed node set. Select a node in one layer and the same node can be identified everywhere else. That means the system is not a group of separate tools. It is one synchronized environment with multiple expressions of the same source structure.
This hot-linking is why confidence builds quickly. A visible object in the model can be traced directly back to its indexed position, its metadata, its values, and its audit state.
The platform supports complete drill navigation up and down the index channel. Drill down into any branch and the system rebuilds around that node. Drill back upward and the broader hierarchy is restored without breaking continuity. This is not ordinary zoom. It is indexed re-centering.
The key point is that lineage is preserved while scale changes. The user can move into a statement, a schedule, a business unit, or a leaf node, and then move back out again while remaining inside one continuous system.
Time is not lost when the user drills. It remains active at every stage. Whether the observer is at the full-enterprise view or deep inside a subtree, the time dimension stays attached to the visible structure. Historical playback, time trimming, scrubbing, and comparison remain available.
That continuity is essential because it allows the user to inspect not only what a branch is, but how it behaves through time while staying in context.
Once the structure is alive and indexed correctly, the system can move from observation into projection. Historical behavior can be extended forward. Scenarios can be applied. Assumptions can be introduced. The important point is that these projections do not sit outside the model as separate calculations.
They propagate through the hierarchy. Changes at one level can flow into higher roll-ups and adjacent dependent structures, allowing the user to see future effects spatially and structurally, not just numerically.
The system earns trust because it combines structure, verification, navigation, time, and projection in one coherent engine. It preserves the index, validates relationships, exposes the audit path, allows full drill movement, keeps time alive, and projects forward without severing the logic of the hierarchy.
That is why B4_UR_IZ® is more than a visualization. It is a structured observational system in which the user can see the data, test the data, navigate the data, and work forward from the data with confidence.
This section explains the operating mechanics behind the system: how data is structured, how the index controls the hierarchy, how the model converts rows into geometry, how time becomes a navigable dimension, and how the 2D and 3D layers stay hot-linked to the same source structure.
The system begins with structured data. Each record is not treated as an isolated row. It belongs to a hierarchy. In the Virtual Time Machine format, that hierarchy is expressed through an indexed lineage, typically with parent-child paths that tell the system exactly where each node belongs.
Metadata is read first, then the time-series fields. This gives every node an identity, a role, and a position before values are rendered. That is the first difference from conventional spreadsheet software: structure is not inferred later. It is declared up front.
The index is the backbone. It tells the engine which node is the parent, which nodes are children, how deep a branch runs, and how all roll-ups must behave. Once that indexed tree is loaded, the same structure drives every other layer: 2D map, 3D model, drill behavior, tooltips, tables, and timeline states.
That means the visual system is never detached from the source data. When a node is selected, the engine already knows its lineage, depth, descendants, and corresponding table entry.
Once the hierarchy is known, the engine converts it into spatial form. Magnitude becomes scale. Depth becomes ring position or spatial nesting. Branch relationships become visible adjacency and containment. Depending on the active mode, the same indexed structure can be rendered as rings, ROSE variants, radial fields, blocks, or other spatial layouts.
The point is not decoration. The point is to turn structural relationships into something the eye can read immediately. Concentration, imbalance, dependency, and scale become visible instead of being buried in row logic.
The engine does not treat time as a passive column. Each period becomes part of a navigable sequence. The model can display a full span, a trimmed span, or a scrubbed interval. This is what makes the platform a Virtual Time Machine rather than a static visualization.
Instead of comparing disconnected snapshots, the observer sees how the same structure changes through time. Growth, compression, reversal, volatility, and persistence become observable as behavior.
The 2D index, the 3D model, and the table view are all tied to the same indexed node set. Select a node in one layer and the engine can identify and surface the same node everywhere else. That is the practical power of the system. You are not moving between separate tools. You are moving between synchronized expressions of the same underlying structure.
This is why the platform supports fast corroboration. A visible object in the model can be traced directly back to its table entry, metadata, and timeline values.
When the user drills into a branch, the engine does not merely zoom. It re-centers the model around the selected node while preserving lineage and time. That lets the observer isolate a statement, a business unit, a district, or any other subtree without losing structural truth.
Large systems become workable because the engine can shift scale while keeping the hierarchy intact.
The value of the system lies in compressing cognition. Traditional analytics make the user reconstruct structure mentally from rows, filters, and reports. This engine does the opposite. It preserves the hierarchy, converts it into geometry, keeps it indexed, and lets time be navigated directly.
That is how the system moves from flat reporting into observation. It does not simply show data. It shows how the system itself is built and how that system behaves through time.
Understanding the Rose Model
A self-contained visual guide loaded directly into this page so the new tab works as a complete section alongside Basic 101, Getting Started, Info, and How It Works.
Info preserves the current brief in its existing voice and structure, while Basic 101 serves as the fuller explanatory framework immediately suitable for orientation, briefing, and conceptual grounding.
The purpose of this split is to give the viewer two synchronized entry points: a concise corporate-facing overview and a deeper system-level explanation.
Data Cocoon LLC, trading as B4_UR_IZ® (Before Your Eyes), is focused on the redefinition of visual data representation — to innovate and disrupt the current status quo in big data analysis and interpretation.
Our mission is to transform data so that the human brain can perceive and visually experience information in a fundamentally new way. By stimulating the primary visual cortex through three-dimensional spatial rendering, B4_UR_IZ® enables users to understand and interpret complex structures with a speed and clarity that no table, chart, or dashboard can match.
Founded by Martin Kagan in 2013, with conceptual development beginning in 2010. Drawing on 35 years of experience in numerical calibration and multi-dimensional financial analysis, and a foundational background in data processing and data controls from his Commerce degree, Kagan developed the concept through multiple iterations over fifteen years — private, patented, and largely unseen.
In April 2024, with the arrival of capable AI development tools, Kagan made a deliberate decision: start over from zero and rewrite the entire system from the ground up — with no prior software engineering background — learning entirely through AI collaboration. The first complete rebuild took three months. The next iteration took three weeks. Then one week. Then daily iteration cycles, often multiple times per day. Hundreds of iterations later, the system you see here is the result.
B4_UR_IZ® is a new category of analytical instrument. Privately funded. Patented structural method. Fully browser-based — every capability runs client-side, no back-end infrastructure of any kind.
The tools used to analyse complex data were designed in the 1980s. The spreadsheet is forty years old. The terminal is thirty. The BI dashboard is fifteen. None were designed for the structural complexity of modern data — financial, operational, scientific, or organisational. They all share the same fundamental failure: they treat hierarchical, multi-period, multi-dimensional data as flat rows and columns.
Financial data is fundamentally hierarchical. Revenue breaks into segments. Segments break into product lines. Product lines carry cost structures that roll into gross margins. Balance sheets nest current assets inside total assets. SEC note schedules — warranty reserves, deferred revenue, debt structures — each carry their own sub-hierarchies. None of the current ecosystem shows you the shape. Excel, Bloomberg, Tableau, Power BI, Capital IQ, FactSet, Refinitiv — all present hierarchical data as tables that collapse and expand. The structure remains invisible.
B4_UR_IZ® taps an innate human capability: spatial cognition. The platform transforms structured data into spatial visual frameworks where hierarchy, magnitude, and time become directly observable. Magnitude becomes spatial scale. Hierarchical relationships become visible containment structures. Time becomes a navigable dimension rather than a passive column of dates.
The Virtual Time Machine (VTM) — the B4_UR_IZ® analytical engine — maps every data node onto a polar coordinate system: concentric rings radiating outward by depth level, angular partitions by sibling position, height by magnitude. The patented infographic renderings are self-explanatory after the briefest initiation period. Unlike elementary tools — bar charts, pie charts, scatter plots, Venn diagrams — B4_UR_IZ® systematically renders multiple structural perspectives simultaneously, in a quick flash of images. The system breathes life into once-inert reams of data.
The Virtual Time Machine is not a financial-only tool. Any structured, hierarchical dataset with a time dimension becomes navigable spatial architecture. The system auto-detects node types, calibrates colour coding and ring depth, and renders the structure through a standard CSV or JSON template import — simple in nature, extraordinarily powerful in output:
| Data Domain | What VTM Renders | What Becomes Visible |
|---|---|---|
| Financial Statements | Income Statement, Balance Sheet, Cash Flow, Equity Changes, 17 SEC Note Schedules — full GAAP/IFRS hierarchy | Where value concentrates, how cost structures cascade, margin evolution across 19+ years |
| KPI Evaluation | Any KPI tree — operational, financial, strategic — with parent-child aggregation across periods | Which KPIs drive which outcomes, where performance diverges from target, period-by-period drift |
| Call Centre & Operations | Agent performance, call volume, resolution rates, escalation trees by region and team | Structural bottlenecks, efficiency concentration, temporal performance patterns across shifts and seasons |
| Survey & Statistical Data | Response hierarchies, demographic breakdowns, longitudinal cohort tracking over time | Structural bias in response patterns, evolution of sentiment across periods, sub-group divergence |
| Organisational Mapping | Headcount, cost, and output hierarchies across departments, divisions, and geographies | Where cost concentrates relative to output, structural redundancy, span-of-control visualisation |
| Global Index Data | Country, sector, and instrument hierarchies — accumulated global statistical datasets over decades | Cross-country structural comparison, sector-level concentration, multi-decade trend trajectories |
| Scientific & Research Data | Any experiment or study with hierarchical variable relationships over time | Structural relationships between variables that flat statistical output cannot reveal |
| Morgan Stanley (Embedded) | 325 nodes, 19 annual periods 2006–2024 — built directly into the current v19 system | Live demonstration: complete financial architecture of a Tier 1 global investment bank |
The current production build contains the following modules, all operating within a single session with no external data transmission. Every figure below is live — not a mockup, not a prototype:
| Module | Capability |
|---|---|
| 3D Studio | Five distinct 3D render modes. Complete financial or operational hierarchy rendered as a navigable spatial model. Orbit, zoom, drill into sectors, isolate subtrees. Scrub 26+ historical periods. Ring-level isolation: toggle any structural depth layer independently across all render modes. |
| DuPont 3D | The only system that renders DuPont decomposition as an interactive 3D structure. ROE apex at depth 1. Profit Margin, Asset Turnover, Equity Multiplier at depth 2. Raw components at depth 3. A widening margin cone means improving profitability — visible at a glance, not decoded from a formula. |
| ROI Analysis 3D | 11 return metrics per period: ROIC, ROA, ROE, Cash ROIC, Gross ROA, Operating Margin, EBITDA Margin, Asset Turnover, Capital Efficiency, Debt-to-Capital. ROIC vs WACC value creation spread: positive = value creation, negative = value destruction. |
| Valuation Engine | 8 methodologies: DCF (adjustable WACC, terminal growth, projection period), Appraisal-Based, Sum-of-Parts by operating segment, Returns Analysis, Comparable Multiples (P/E, P/B, EV/EBITDA, P/S, P/FCF), DuPont Decomposition, and a 10,000-run Monte Carlo simulation generating fair value confidence intervals. |
| Observatory | Structural analysis layer: which nodes grow fastest, where value concentrates, CAGR computed across every branch of the hierarchy, volatility distribution across the full tree. Multi-pass bottom-up aggregation. |
| Forecast Engine | 5 projection methodologies auto-selected per node based on statistical profile: HIGH_GROWTH_DECAY, STABLE_GROWTH, RECOVERY_TRAJECTORY, MEAN_REVERSION, CAGR_DAMPENED. Ratio-aware forecasting for profitability, efficiency, and leverage ratios with distinct reversion and range-boundary logic per ratio type. |
| Schedule Engine | 5-schedule double-entry ledger: Base Data, Journal Entry Deltas, Projections, What-If Percentage Adjustments, Computed Result. Balance verification runs automatically. Deep-clone backup and restore across all five schedules. |
| 2D Dashboard | Full tabular analytical view. Bidirectional with 3D: click a node in 3D and the table highlights. Click a table row and the 3D model focuses. Multi-column sort, filter by type, depth-level drill-down, share price integration. |
| Compare Mode | Two companies simultaneously in split-screen with synchronised camera and time. Cross-company narrative report generated automatically: revenue CAGR, margin, ROE, leverage, asset trajectory. |
| NEWVIEW Orbital | 7 concentric orbital rings by depth. Per-depth magnitude normalisation — all structural levels simultaneously legible regardless of raw value scale. Bezier parent-child arcs in 3D space. Temporal history ghosts. Full hover tooltips, click selection, double-click drill. 3-stage inspection exit. |
| Audit Trail | Blockchain-stamped record of every data modification. Traceable, verifiable, non-repudiable. Every adjustment logged with timestamp and structural context. |
| Import Engine | Standard CSV or JSON template import. Any structured hierarchical dataset with a time dimension loads and renders automatically. The system auto-detects node types and calibrates the spatial model accordingly. |
| Pre-Evaluation Analytics | On dataset load, the system runs a structural pre-evaluation before the user makes a single interaction. It identifies concentration anomalies, flags outlier nodes, surfaces structural patterns, and delivers guidance points — telling you where and what to look for before you know to ask. The analytics lead the investigation; the user follows the signal. |
Apple. Microsoft. Nvidia. Alphabet. Amazon. Meta. Tesla. The dataset in active development. These companies represent the most structurally complex publicly reported financial architectures in existence — deep hierarchies, 15–20 years of reporting, hundreds of interconnected metrics, trillions in cross-period variance. VTM renders all of it simultaneously, spatially, in real time.
The Tesla proof-of-concept — the primary reference dataset — contains 234 nodes across Income Statement, Balance Sheet, Cash Flow, Equity Changes, 17 SEC Note Schedules, 3 operating segments, and 15 financial KPIs, spanning 26 historical periods plus 10 projected periods. If it works on Tesla's structural complexity, it works on anything.
This trajectory is itself a demonstration of what AI-assisted development makes possible. The question of how a single non-engineer built a system of this capability in under twelve months is not separate from the story of the system — it is part of it. The tools that made VTM buildable are the same tools that are redefining what can be built, by whom, and how fast.
| Capability | Why It Matters |
|---|---|
| Structural fidelity | The 3D model preserves actual GAAP/IFRS hierarchical relationships. You see the tree as the tree — not a proxy, not an approximation. |
| Bidirectional navigation | Click a 3D node: the table highlights. Click a table row: the 3D model focuses. Every analytical finding is spatially addressable. |
| DuPont + ROI in 3D | The only system rendering DuPont decomposition and ROI analysis as interactive navigable 3D structures. Margin, turnover, and leverage as spatial dimensions — not three numbers in a formula. |
| 10,000-run Monte Carlo | WACC, terminal growth, and projected cash flows randomised across normal distributions. Fair value confidence intervals generated in seconds. |
| Five-schedule ledger | Double-entry accounting with balance verification. Base, journal entries, projections, what-if, and computed result — all in one session. |
| Dataset-agnostic import | CSV or JSON template. Any structured hierarchical dataset with a time dimension. The system configures itself automatically. |
| Universal access | No server. No subscription. No installation. A regulator in Lagos, a student in Jakarta, and a fund manager in Zurich have identical analytical access. |
| Phase | Objective | Status |
|---|---|---|
| Mag-7 Full Suite | All 7 tech giants — cross-company narrative AI, synchronised comparative analysis at full depth | ● In Progress |
| NEWVIEW v2 | Animated time-sweep morph. Branch shell envelopes. Enhanced temporal trace density. | ● In Progress |
| Observatory Mode | Macro layer: CAGR mapping, volatility distribution, growth concentration across full hierarchy | ◆ Planned |
| Live Data Feed | SEC EDGAR and institutional data integration — real-time node rendering as filings release | ◆ Planned |
| Sector Orbital | All Mag-7 simultaneously on one concentric field. Sector-level spatial index with cross-company arcs. | ◆ Planned |
| Valuation Engine v2 | DCF / DDM / Comparable multiples rendered as 3D objects overlaid on the hierarchy | ◆ Planned |
| VR Mode | WebXR rendering — walk through a company's financial architecture at full human scale | ◈ Exploring |
| Education Layer | Guided annotation mode — anchored labels, structured walkthrough paths, student interaction logging | ◈ Exploring |
This system was built by Martin Kagan — a finance professional with 35 years of numerical calibration experience and a foundational training in data processing and data controls. The concept was first developed around 2010. Fifteen years of iteration followed, much of it unseen. Then in April 2024, AI arrived as a genuine development partner — and everything accelerated.
Starting from zero software engineering knowledge, Kagan rebuilt the entire system from scratch using AI as the co-author of every line of code. The first rebuild: three months. The next: three weeks. Then one week. Then daily — multiple iterations per day, hundreds of cycles of refinement. What you are looking at is the product of that process. No corporate team. No external engineering resource. One person, one concept, and the tools that finally made it buildable.
The system is fully browser-based. There is no back end. No server. No infrastructure. Everything — advanced analytics, valuation models, Monte Carlo simulation, DuPont decomposition, five-schedule ledger, blockchain audit trail — runs entirely in the browser. And the analytical layer does not wait for you to explore. It pre-evaluates the dataset on load, identifies structural anomalies, flags concentration points, and surfaces guidance before you have made a single click. The system tells you where to look before you know to ask.
B4_UR_IZ® is a private enterprise. We are not recruiting staff. What this system requires is something more valuable: people who are willing to take the time to genuinely understand what they are looking at — and then help carry the idea forward. The spreadsheet had a good run. It is time to see the shape of the data.
The site carries over 80 videos covering the full range of system capabilities and data domains:
Balance Sheets · Income Statements · Global Index Data · Organisational Mapping
DuPont Decomposition · Monte Carlo Stress Testing · The Magnificent Seven
If this is relevant to your work or study — share it. The concept is new. It benefits from the widest possible audience.
We are not seeking staff or partners at this stage. We are looking for people who will take the time to understand the system — and help carry the idea forward.
