MatLogica’s advanced analytics increased speedups for a Tier 2 European bank by 15-20x

MatLogica is a UK-based early-stage deep-tech company with revolutionary technology in the fields of high-performance computing and automatic adjoint differentiation. Founded by a quant with 17+ hands-on experience and a mathematician, MatLogica developed a framework that uses a fundamentally different approach to working with a CPU allowing us to extract the highest level of performance from modern hardware. Our framework is a perfect fit for finance, medical research and diagnosis, space research, geophysics, weather modelling, and more.

Register Interest

With the launch of their product line, Matlogica has put together a case study with a Tier 2 European Bank

Client: Tier 2 European Bank

Client problem: 

To compute risk, the client’s proprietary quant system employed a combination of Bump & Revalue and manual adjoint differentiation. Despite substantial expenditure, the risk system struggled to provide risk reports on time for the next trading day.

Solution and results:

Analytics powered by MatLogica opened revenue streams, reduced infrastructure costs, and improved risk management. When compared to their old manual adjoint differentiation paired with Bump & Revalue, they obtained 15-20x speedups.

Overnight portfolio risk computation was reduced from 8+ hours to 2 hours, while intraday risk calculation was reduced from 30+ minutes to a few minutes. By reducing difficult-to-maintain elements, the modelling quant team’s pace rose threefold, allowing speedier time-to-market for new features.

Register Interest