Dune grass overlooking a wide North Sea beach near The Hague, with WWII-era bunkers along the dune ridge

Quantology

Advisory and tooling at the intersection of model risk and AI, for banks and insurers.

Golden sunset reflecting over the North Sea

“The same seniority and methodology as the Big 4, delivered faster and at lower cost through self-built AI tooling — without compromising on quality, explainability, or fit with existing governance requirements.”

Services

Three fixed-price engagements, from a low-threshold first introduction to deep, senior-led model work.

Model Inventory Tool

A structured overview of all quantitative models within the organization (AIRB, IFRS 9, Solvency II and related), including risk classification, ownership, regulatory linkage, and documentation status.

Problem it solves

Model risk management functions often lack an up-to-date, complete, or standardized overview of their model landscape — resulting in governance gaps, missed review deadlines, and cumbersome reporting to the regulator.

Approach

Base: generic, proven taxonomy (model name, owner, risk tier, regulatory driver, review date, documentation status)
Tailored to the client's specific model classification, reporting format, and supervisory framework (DNB / ECB / group level)
Delivered as a working prototype, immediately usable as a governance starting point

Fixed price, short-term (weeks, not months)

Entry engagement / first introduction — low-threshold, quick demonstrable result

Open the model inventory tool

Model Monitoring Pipeline

An automated model performance monitoring solution: PSI, backtesting, override tracking, and drift detection, connected to the client's data sources.

Problem it solves

Model monitoring at many institutions is still manual or ad hoc, resulting in late detection of performance deterioration and high repeat costs per reporting period.

Approach

Core engine (PSI, backtesting, override tracking, alerts) pre-built and tested
Data source connection built as a swappable module — configured to the client's specific data source and model types
Reporting output aligned with internal and supervisory requirements (IFRS 9 / AIRB)

Fixed price, per model (family)

Institutions that already want standardized monitoring but haven't automated it yet

Model Development & Validation Acceleration

Accelerated model development, re-documentation, or second-line validation of AIRB, IFRS 9, and Solvency II models, supported by ModelMaker — self-built AI tooling that helps generate, document, and pre-validate models.

Problem it solves

Traditional model development and validation are time-consuming and expensive, with documentation burdens that are often disconnected from the actual modeling process. This slows implementation and raises the cost of external expertise.

Approach

15 years of senior experience (Lloyds Bank, Achmea, Deloitte) as the domain foundation
ModelMaker accelerates model building, documentation, and pre-validation — an efficiency advantage, not a standalone product
Output remains traditional and explainable, fitting within existing governance and supervisory frameworks

Fixed price per model or model track

Institutions needing senior expertise in model development or validation, faster and more competitively priced than traditional consultancy

Approach

Quantology is built on 15 years of senior model risk experience across Lloyds Bank, Achmea, and Deloitte — spanning model development, validation, and governance for AIRB, IFRS 9, and Solvency II frameworks.

That experience is paired with ModelMaker, self-built AI tooling that accelerates model building, documentation, and pre-validation. It's an efficiency advantage, not a standalone product: the output remains traditional and explainable, fitting within existing governance and supervisory frameworks.

A quiet dune pond surrounded by reeds and dune grass near The Hague

Contact

Considering a low-threshold first engagement, or ready to discuss a specific model track? Send a message and I'll respond directly — or reach me at m.malkus@gmail.com.