SEADS: Empowering Shell's Data Revolution

Overview
Together with Shell Trading & Supply, we built Shell’s first data marketplace, SEADS, leveraging a data mesh paradigm. The platform provided democratised data access paired with seamless discovery, utilisation and reusability.
KEy Metrics
12x
faster discovery of new applications and services
30
min
for service registration from 10+ hours
70
+
new services created using our templates in 2 months
11
product teams developed application using our patterns
The Challenge
Shell Trading & Supply business was defined by data silos, underutilised information assets, high development costs, lack of a clear governance structure, and slow innovation pace. These led to duplicated efforts and inconsistencies in data service development, thereby complicating the reuse of existing models. As such, the client faced multiple challenges in effectively and efficiently discovering, understanding, trusting, and utilising high-quality data.
Our Solution
The SEADS data marketplace was designed and developed to transform the Shell Trading & Supply’s data into assets that can be easily accessed and leveraged by their employees. Leveraging a data mesh paradigm, it consists of three layers, working together to reduce the friction of starting a data-heavy project:

GLASS
with its own SDK

Serves as the marketplace’s entry point and provides the user with global navigation and search widget that ensures effortless discovery of available data. Its user-centric design ensures a uniform appearance and experience of products across Trading & Supply, streamlined onboarding, and minimised learning curve for users.

DATA LAKE
with its own Python SDK

Unifies diverse data sources offering a consolidated perspective of Shell's data assets. It does so through an abstraction layer, exposed via an API, and leverages tools such as Databricks, Purview, Apache Superset, Azure APIM, and a wide variety of day-to-day cloud services.

DESIGN MODULE
with its own React SDK

Design Module serves as a library of reusable design patterns, code library and interactive storybook.

Key Focus Areas

Robust Architecture

We leveraged a data mesh paradigm to ensure the firm can easily utilise a reliable ecosystem of interconnected data products. This enables smooth exploration of the extensive data environment.

Empowering Designers & Developers

We provided a comprehensive component library, which made it possible for Shell's design and development teams to make use of reusable patterns. This approach accelerated the process and enhanced the effectiveness of their creations.

API Exposure & Acceleration

We developed a translation layer to expose APIs. This was further boosted by an API accelerator and a low-code platform, fostering collaboration and accelerating the creation of innovative applications and services.

Data Discovery & Reusability

SEADS enables a clear overview and access to historical data, ensuring reduction of duplicated efforts and inconsistencies in data service development.
Deep dive
Recognising the divergence between big data technology stacks and operational platforms, SEADS integrates advanced technological solutions to bridge this gap, facilitating seamless interoperability and interconnectivity between distinct sets of infrastructure. It does so by leveraging the latest in big data technologies, enabling domain teams to deploy their services using cutting-edge tools.

We use an array tools to ensure the platform's versatility and capability to handle diverse technological ecosystems. For example, we use containers such as Docker (for deployment) and Kubernetes (for orchestration), Spark jobs on Databricks clusters for their pipeline code, and the serving capabilities of MLFlow so that we can integrate AI models with our APIs. This enables us to facilitate the access from ML clusters to the datasets in a lively manner, accelerating the process of data gathering for AI.
the impact

Enhanced Data Quality & Utilisation

By treating data as a product, SEADS ensures that information assets are no longer underutilised, promoting a culture of data excellence and innovation.

Reduced
Operational Friction

The platform significantly reduces the friction and costs associated with discovering, understanding, and trusting data, enabling Shell to make more data-driven informed decisions.

Interoperability & Connectivity

SEADS addresses the challenges of integrating distinct sets of infrastructure, offering a unified platform that caters to the diverse needs of big data and operational technologies.

Customer Satisfaction

With a focus on treating data consumers as valued customers, SEADS enhances user experience, reduces service registration with 95% (from 10 hours to 30 minutes).

Federated Governance

The operating model is based on a federated decision-making and accountability structure, with a team of domain representatives, data platform, and subject matter experts—legal, compliance, security.
conclusion
The SEADS platform represents a paradigm shift in how organisations approach their data, offering a comprehensive solution to the challenges of data quality and silos. By fostering a culture of analytical excellence, SEADS empowers businesses to unlock the full potential of their data, driving innovation and operational excellence.
FESTINA LENTE

Curious to learn more?

Reach out to us!