IG Group's New Data Model Wins Google Backing in AI Readiness Race

Most large organizations claim to be data-driven, yet their teams still spend a majority of time cleaning and locating data rather than building models or AI products. IG Group has attempted to solve this with a new "extended medallion architecture," which gained significant endorsement from Google, resulting in a co-authored white paper and a customer success story.

The core issue is a familiar one in data management. Traditional platforms often use a medallion pattern: bronze for raw data, silver for cleaned data, and gold for business-ready outputs. For FX and retail trading, speed and consistency are paramount. IG's new model aims to provide its FX, CFD, and crypto teams with cleaner data faster, enabling quicker feature launches, reporting, and risk tool deployment to react to market movements.

For clients, this should translate to more reliable information across platforms and a faster rollout of new, AI-driven tools. Essentially, IG is fixing its data infrastructure to deliver clearer, more dependable information to traders sooner. The extended architecture maintains central control over core data layers while allowing business teams like risk and finance to build their own "domain gold" projects without overloading central engineers or altering the core model.

The design is platform-agnostic but is currently running live on Google Cloud using services like BigQuery. A senior Google field engineer stress-tested the architecture before Google agreed to publicly endorse it with detailed documentation. For IG, the model seeks to reduce data preparation burdens and accelerate product development. For the market, it presents a template for balancing governance with agility as AI workloads increase.

While many large financial institutions are quietly rebuilding data platforms for AI, they rarely share full blueprints. IG's move is notable because Google has validated and published its architecture as a reference model. For a retail and FX broker of IG's scale, such recognition from a major cloud provider is exceptional.

In a contrasting development, IG Securities (IG Japan) recently disclosed a data handling lapse that potentially exposed the personal information of up to 162,879 clients internally. The firm stated there is no evidence of an external breach, attributing the incident to contractor oversight and weak access controls.