What exactly does the AI do, and what does it not do?

It turns a plain-language description into a scenario draft. It does not value anything, decide anything or save anything without explicit validation. Valuation and sensitivity calculations are deterministic and use no language model.

What does a demo look like?

30 to 45 minutes, no commitment, starting from a sample portfolio, or from a scenario representative of your context if you describe it beforehand.

How does a first deployment work?

First implementations are scoped pilots: scope, portfolios and duration fixed upfront, with a weekly progress review.

Available

Rate and spread scenarios, down to the position.

Compose shocks layer by layer, value the portfolio and read differences line by line.

Request a demo See the product

Multi-layer rate and spread scenario builder in Fixed Income

01Scenario

Build the scenario.

From the assistant to the versioned library, without leaving the application.

The assistant drafts, the analyst validates

You describe a shock in English or French and the assistant produces a scenario draft. Nothing is created before validation, and the resulting scenario is an editable object like any other: each layer can be reworked by hand.

Assistant: a shock described in plain language, turned into a scenario draft — Draft produced by the assistant, before validation. Shock layers are listed and editable one by one.

Shocks are built layer by layer

Each layer carries its type, axis and magnitude. They stack in the scenario, stay visible and can be edited one by one. The impact preview shows the curve before and after.

Multi-layer rate and spread scenario composition — Scenario under construction: stacked shock layers, with curve impact preview.

The library keeps the versions

Drafts, simulations and published versions live in the same place. A scenario from last exercise can be duplicated and evolved without starting over and without losing the original version.

Scenario library: drafts, simulations and published versions — Scenario library, with published versions and drafts in progress.

02Deliverables

What a run produces

Valuation, differences, duration and sensitivities, line by line and in summary.

The assumption trail follows in the report.

Detailed results: line-by-line valuation under baseline or scenario — Detailed results: valuation, differences, duration and sensitivities, one row per position.

Run summary: key indicators per scenario — Run summary: impact per scenario, contributions per segment, sensitivities.

Download a sample report (PDF)Produced by the application on a test portfolio.

03Coverage

What Fixed Income covers.

Shocks, axes, calculations and data exchanges.

Functional coverage
Shock scope	Rates and credit spreads. Per layer and per maturity segment: parallel, scale and twist.
Application axes	Currency, sector, rating, curve family. Axes combine within a single layer.
Composition	Layers stack within a scenario. Each scenario stays independent.
Calculations	Baseline and scenario valuation, differences, duration and sensitivities, down to the position.
Inputs	Portfolio and market data imported by file. Instruments are linked to curves through their characteristics (sector, currency, etc.).
Outputs	Result tables, KPI summary, run report PDF and assumptions trail, exportable at any time.

04Security

Hosting, access and data handling

Where the data runs and who can access it.

Hosting: Microsoft Azure, European region, with application environments and data separated per client.
Access and logging: Microsoft Entra ID, named accounts, role-based permissions, logging of actions and parameter changes.
Client data: Imported portfolios stay in your environment and do not feed the training of any model.
Documentation: Architecture file, data flows and access matrix: documented and available on request.

05FAQ

Three frequent questions.

What exactly does the AI do, and what does it not do?: It turns a plain-language description into a scenario draft. It does not value anything, decide anything or save anything without explicit validation. Valuation and sensitivity calculations are deterministic and use no language model.
What does a demo look like?: 30 to 45 minutes, no commitment, starting from a sample portfolio, or from a scenario representative of your context if you describe it beforehand.
How does a first deployment work?: First implementations are scoped pilots: scope, portfolios and duration fixed upfront, with a weekly progress review.

Another question? Write to us

See Fixed Income on a case close to yours

Describe the kind of portfolio and scenario you care about and we prepare the demo on that basis.

We reply within one to two business days.

Request a demo Email us directly