Experience

Selected problems and delivery experience

Much of Jaime’s earlier work was carried out inside global financial institutions. The examples below focus on the type of problem, the approach taken and the operational result without disclosing confidential details.

Impairment modelling

Modernising a long-running production process

The situation

A core impairment-modelling process had been maintained for years. Improvements were being requested, but the established team and approach were not able to deliver them within the existing legacy stack and workload.

The approach

Jaime took a different technical approach, redesigning the workload and moving it from SAS to Python and Hadoop. The aim was not simply to translate the existing code, but to rethink how the process used the available data and computing platform.

Result

Runtime reduced from several days to under an hour, creating a more scalable production process.

Explain and sensitivity analysis

Replacing manual, ad hoc analysis with a systematic framework

The situation

Explain and sensitivity analysis had developed independently across models and teams. The work was often manual and ad hoc, making results harder to compare, govern and use consistently in decision-making.

The approach

Jaime introduced systematic, repeatable frameworks across capital, impairment and stress-testing models. The work created a common way to attribute changes, test sensitivities and communicate what was driving the result.

Result

Greater consistency, clarity and governance, with stronger support for technical and commercial decisions.

The common thread

Work that performs under scrutiny

Start with the constraint

Understand what the existing approach can no longer do before choosing a replacement or adding more technology.

Make the result explainable

Build outputs and controls that let the people using the system understand, challenge and trust its results.

Design for production

Treat performance, integration, ownership and operational reliability as part of the solution from the start.

Start a conversation

When the next step is not obvious, let’s talk it through

You do not need a complete brief. An initial conversation can help clarify the problem, the available options and whether Signal Boost is the right fit.

If an NDA would make it easier to talk openly, we’re happy to sign one.

Start a conversation