Cross-source reasoning

We read people
through food.

We surface what a decision is quietly standing on — and where it leads. Traced to real sources, not a crystal ball.

What CusiLabs builds

We build the whole stack.

Cross-source reasoning is a method for reading risk that lives between disciplines — consumer, cultural, market, and beyond — by connecting many sources into one traceable structure. It surfaces contradictions and the consequences of consequences, not predictions and not opinions. From the raw source to the finished read, every layer is ours — not a model with a wrapper. That’s why the reasoning is traceable, and why we can stand up a new graph for a new domain.

  1. 1.0

    Extractors

    We turn unstructured sources into structured, checkable facts.

  2. 2.0

    Ontology

    A proprietary model of how human domains actually relate.

  3. 3.0

    Cross-source graph

    Evidence from many sources, connected into one structure.

  4. 4.0

    Derived mechanisms

    Mechanisms, tensions and patterns no language model could surface on its own.

QC Brain · cross-source graph

One thread through every domain.

QC Brain is a cross-source graph that reads consumer, cultural and market risk through food — every claim traced to its sources. Everyone eats — so food is the one thread that runs through culture, history, economy, society and medicine. These domains converge at the table, every day. QC Brain makes that structure visible, and traceable.

Many domains, one structure.

Evidence

Patterns others miss.

The same cross-source patterns repeat across companies, countries and decades — in the gaps between fields, where no single team is looking.

What one read looks like — a decision, traced to what the graph already held.

Sample read retrospective · illustrative
Decision

Take an infant-formula product into a new market — safe at home, a logical expansion.

What the graph derived
  1. TensionA safe product that leans on the breastfeeding it replaces.
  2. MechanismFree hospital samples start formula; the mother’s own milk stops.
  3. Second orderOnce her supply is gone, it rarely returns.
  4. Third orderThe free supply ends — and now there is no way back.
  5. PatternThe Invisible Substrate — break the cultural ground a decision stands on, and the cost is irreversible.
In the record

Infant illness · an international boycott · a 1981 WHO Code. · WHO Code, 1981

Traced across
  • infant biology
  • economics
  • public health
  • policy

Retrospective — the graph already held the tension. We don’t claim we predicted it.

Request a read

Lenses we apply in a read:

Invisible Substrate Scan

The unspoken cultural ground a decision quietly stands on.

Operational Fragility Scan

Efficiency that holds — until the one condition it depends on moves.

Heritage Risk Review

When extraction quietly erodes the legitimacy it depends on.

Policy Backfire Assessment

Interventions that push the system the opposite way.

Double-Edge Mechanism Review

The same mechanism that helps in one context harms in another.

How it helps

From the decision you’re about to make — to the risk you couldn’t see.

We reduce risk — we don’t sell certainty. Every step is traceable to the evidence behind it.

  1. 01 Your question You’re about to launch a product, set a policy, or change a diet.
  2. 02 Into the graph We connect it to the cross-source graph — every force that bears on it, not just your own data.
  3. 03 Tensions surface Contradictions, hidden dependencies, second- and third-order effects others miss.
  4. 04 A memo A clear, plain-language brief — every claim traced to its real sources.
  5. 05 A better decision You move with the risk made visible. Traceable risk reduction — not a guess.

Who it’s for

When the answer lives between the fields.

A launch. A policy. A reformulation. A market entry. The forces that decide the outcome rarely sit inside any one team’s data.

Investment funds

Before you back a food or consumer bet — the cultural and behavioural risk a commercial due diligence never quite reaches.

Food & CPG companies

Before you launch or reformulate — see why a market will accept it or quietly reject it.

Beverage companies

Before you enter a market — see the meaning and rituals your category data misses.

Food-strategy consultancies

Before you defend a recommendation — back it with a cross-source read your client can audit line by line.

Policymakers & public health

Before a guideline, an intervention or a national programme — see how it could backfire across populations, culture and trust.

Products

The system, applied.

QC Brain

The cross-source graph that reads consumer, cultural and market risk through food.

Cusikay

coming soon

Cusikay is a consumer app for recipes and easy, human eating — built on the same cross-source graph that powers QC Brain. In development; not yet released.

About Cusikay

Born from a kitchen and a passport.

“When something feels natural, you stop seeing it. That is exactly where the structure hides.”

CusiLabs began with one person noticing the same thing everywhere. A trained chef (Le Cordon Bleu, Peru) with a background in engineering and business, he spent years travelling and living across Peru — a crucible of cultures — and then Southeast Asia, India, the Middle East, Russia and Europe.

Everywhere, the same quiet pattern — and the same blind spot: the forces that shape what people eat were never read together. QC Brain is the attempt to read them together — and the first of the graphs we can build.

Request a read

See what your decision is standing on.

Name a decision you’re weighing. We’ll run a traceable read and show you the risk your own data can’t.

One decision at a time.