Veritas Per Se. ⟩

The Fractal Nature of Truth

A fractal is a structure that repeats its fundamental pattern at every scale of observation. Truth exhibits this same property: examine a genuine dataset at the macro level or the micro level, and the underlying signal remains self-consistent.

  • Recursive consistency. Zoom in as far as you wish — the core pattern does not contradict itself. Mandelbrot (1982) demonstrated that financial markets, far from being Gaussian, exhibit fractal self-similarity.
  • Infinite depth. Truth is not a surface. Each layer of investigation reveals new detail without invalidating the layer above it.
  • Organic emergence. Truth is not engineered; it arises from the conditions that produced it — much like a coastline or a branching tree.
  • Dynamic equilibrium. Truth is not static. The coherence-entropy conservation principle predicts that any localised increase in structural coherence must be balanced by a redistribution across adjacent scales — truth reorganises itself but never contradicts itself.

Ref: B. Mandelbrot, The Fractal Geometry of Nature (1982); E. Peters, Fractal Market Analysis (1994); Coherence-Entropy Conservation (STAT Framework)

Truth has a shape. 3S ⟩

The 3S Framework — Measurable Architecture

If truth has structure, then it can be observed, measured, and tested. T2SAIM's operational premise rests on three forensic axes — the 3S:

  • Shape — The morphology of a dataset reveals whether it grew organically or was deliberately arranged. Genuine distributions have characteristic shapes; fabricated ones do not.
  • Structure — Relationships between data points form networks whose topology is diagnostic. A fraud ring leaves a structural fingerprint distinct from legitimate commerce.
  • Scale — Fractal self-similarity means that authentic patterns repeat across scales of observation. Mandelbrot's scale invariance is the litmus test: if a signal breaks down when you zoom in, the signal was manufactured.

The 3S axes are operationally measured through the Conspiracy Index (KE) matrix, which tests three structural columns of any narrative: chronological consistency (time), relational naturalness (network topology), and linguistic entropy (synthetic language score). If a narrative fractures on any column, the structure is flagged as unsafe.

Ref: Scale invariance (Mandelbrot, 1982); Network topology (Barabási, 2002); Conspiracy Index — KE Matrix (T2SAIM)

Beyond the PsyOp. ⟩

The Geometry of Deception

Where truth is fractal, deception is Euclidean: straight lines, calculated angles, artificial forms. A lie requires an architect. It is designed to create a specific perspective for a specific audience.

  • Engineered construction. Deception cannot emerge spontaneously. It requires deliberate assembly — a blueprint, a narrative arc, a coordinated set of actors.
  • Rigid boundaries. Fabricated structures are closed systems. They cannot tolerate infinite-depth scrutiny, because at some resolution the seams become visible.
  • Forced perspective. Like a geometric illusion, deception only appears coherent from a single vantage point. Change the angle — and the void behind the façade is exposed.
  • The paradox of perfection. Organic processes contain entropy — natural imperfection. If a dataset, a financial statement, or a narrative appears excessively orderly, that very regularity is diagnostic. Artificial perfection is the signature of manufactured data; T2SAIM flags low entropy as a synthetic trace.

Ref: Econophysics of deception networks; Information entropy in forensic analysis; Graph theory (Erdős–Rényi)

We verify independently. ⟩

Independent Forensic Verification

T2SAIM separates analysis into two cognitive stages. Stage 1 scans the full data field for statistical, structural, and behavioural irregularities. Stage 2 subjects each candidate to strict evidentiary protocol.

Every finding is classified: verified, assumed, or unverifiable. No advocacy. No narrative. Only the epistemic status of each claim, traced to its source.

All forensic findings are sealed with cryptographic timestamps (SHA-256) and mapped against the Triple-Witness Seal protocol to achieve Daubert-compliant evidentiary standards. Every claim is traceable from raw data to final classification.

You decide with clarity. ⟩

Clarity as Deliverable

T2SAIM does not produce recommendations. It produces the evidentiary conditions under which a competent decision-maker can act with confidence. The decision remains yours — the clarity is ours to provide.

3S · 3D
Detect · Defend · Decide
We are the Verity.
Explore the method → Investor relations →
Scroll to explore

Most analytical tools assume the data is honest — noisy perhaps, incomplete certainly, but not strategically deceptive. That assumption fails the moment a market is being manipulated, a document is being weaponised, or an institution is being misled by its own internal feedback loops.

T2SAIM was built for the cases where the standard assumption breaks. Where the noise is not random but engineered. Where the question is not what does the data say but what is the data trying to make us believe.

A two-stage approach to contested information

T2SAIM separates analysis into two distinct cognitive operations.

01

Broad Anomaly Sweep

The full data field is scanned for statistical, structural, and behavioural irregularities. Stage 1 is deliberately permissive. False positives are expected and welcome.

02

Forensic Interrogation

Each candidate is examined under strict evidentiary protocol. Every claim is sorted: verified, assumed, unverifiable. Only findings that survive this filter are reported.

Veritas Per Se — the epistemic backbone

Verified

Established with primary evidence

Assumed

Working assumption with stated provenance

Unverifiable

Open question, named explicitly

Designed for high-consequence domains

Three application verticals where the cost of false confidence is highest.

Fraud Analytics

Detect coordinated manipulation in financial data, payments, and trading flows that conventional pipelines miss because the adversary adapts faster than the model.

Forensic Intelligence

Apply structured anomaly detection to geopolitical, regulatory, and open-source intelligence streams where both the signal and the noise are strategically authored.

The cost of false confidence is rising

Generative systems have made plausible fabrication cheap. Adversarial actors have learned to craft data, documents, and narratives that pass automated checks. Regulators — particularly under the EU AI Act and emerging UK frameworks — now expect institutions to demonstrate not just outputs but defensible reasoning.

T2SAIM is built for that requirement. Every finding it produces is traceable to its epistemic source. Every claim is tagged. Every uncertainty is named.

The work behind the method

📄

Technical White Paper — v9.6

The T2SAIM technical corpus comprises the methodology specification, the formal claim set (19 independent + 2 dependent), and the application-vertical extensions. A redacted summary is available on request.

Request access →

Foundational references

A method needs a methodologist

T2SAIM was developed by Tarkan Bulan, an independent researcher and analyst with over a decade of work spanning epistemic security, forensic linguistics, macroeconomic intelligence, and the philosophy of evidence.

The framework emerged from a recurring observation across very different domains: that the same class of error kept producing the same class of failure, and that the error was not technical but epistemic.

Building beyond the founder

A methodology that lives only in one head is not a methodology — it is a habit. The roadmap for T2SAIM Ltd is the institutionalisation of the framework: documentation, formal training pathways, and a senior team capable of running engagements without founder dependency.

T2SAIM Ltd

A UK-registered private limited company, incorporated to develop, license, and deploy the T2SAIM methodology across regulated and high-stakes domains.

Legal nameT2SAIM Ltd
JurisdictionEngland and Wales
Registered office[To be confirmed]

Governance and conduct

Get in touch

For methodology enquiries, pilot conversations, research access, or investor relations.