What is SDP?
The Standard Data Platform (SDP) is the whole-of-government data infrastructure that lets any Singapore government agency stand up a secure, governed, cloud-native data lakehouse in minutes, not months.
Today, every agency that wants to do serious analytics or ML has to reinvent the same wheel: procure a platform, wire up networking, pass security review, set up governance, and figure out how to share data with other agencies. SDP does that once, centrally, and correctly, and then serves it all out through a single self-service portal. Think of it as a multi-tenant platform-as-a-service for government data, enabled and delivered through the SDP portal.
The portal is how the whole platform reaches agencies. An agency registers on it, picks a tier, and gets a fully provisioned workspace served out automatically. Through the portal an agency can choose from best-in-class engines (Databricks, Snowflake, AWS SageMaker Unified Studio, and soon Microsoft Fabric), all running inside Singapore's Government Commercial Cloud (GCC) on AWS and Azure. Agencies publish datasets to a central catalog (DataHive) and share them across platforms using open table formats such as Iceberg and Delta Sharing. Everything is secure by default: private networking, no public internet data paths, audit logging to the government SOC, and AI access locked to a single approved gateway.
This is real, in-flight infrastructure serving real agencies, not a lab project. Government data is fragmented across agencies and platforms, and SDP is the plumbing that makes it safe and easy to onboard, govern, and share, which is what unlocks whole-of-government analytics and AI. When it works, an agency spins up a compliant data platform in under a week instead of running a six-month procurement-and-security marathon. The team's north star is simple: the number of agencies actually running on SDP. Everything else, from the automation to the security controls to the catalog, serves that number.
How the SDP portal serves out SDP setups
The portal is how an agency gets its SDP setup: a ready-to-use, governed data platform of its own. Without it, an agency waits days while engineers set everything up by hand. With it, an agency signs in, requests what it needs, and the platform provisions it automatically, with live status updates the whole way. The goal is to go from request to a working setup in under 30 minutes instead of days.
Behind that one button is the part you would help build: the automation that turns a simple request into a fully provisioned, secured, and connected data platform. A request flows through sign-in and access checks, into an API, and out to a provisioning pipeline that stands up the agency's environment as code and wires up its data catalog, networking, and security, with no manual changes to production. Turning a form submission into a live, governed data platform is the core of what the team builds.
What we are building next
1. The foundation is in place; the interesting work now is growing the platform outward. These are open, greenfield problems where you help shape the answer, not maintain a finished one:
2. Expanding the engines. SDP started with Databricks and Snowflake. We are extending it to AWS SageMaker Unified Studio and Azure Fabric, so agencies can run analytics, ML, and AI on whichever platform fits their work, all provisioned through the same self-service flow.
3. Streamlining data sharing across SDPs. We are exploring and designing how one agency can discover and consume another's governed data, brokered through DataHive, our central data catalog, without bespoke, one-off integrations. This is a design-heavy problem: you would help define how data sharing works across the whole of government.
4. Validating that agencies conform to SDP requirements. As more agencies join, we need to check, automatically, that each one's setup actually meets SDP's standards for security, configuration, and governance, and stays that way over time. We are building the conformance checks and controls that catch drift and non-compliance before it becomes a risk.
5. Raising the security posture. We are hardening SDP to handle even higher data sensitivities, opening the platform to workloads that cannot run on shared cloud infrastructure today.
6. Enhancing common WOG infrastructure. We are improving the shared whole-of-government services SDP depends on, so every agency on the platform can reach common AI and data services through one governed, compliant path.
