Architecture

Your School Already Runs on Google Cloud

Systemic Analysis
Your School Already Runs on Google Cloud

Every school using Google Workspace for Education is already a tenant on one of the most powerful cloud platforms on the planet. The same data centres. The same APIs. The same infrastructure that runs Google Search, YouTube, and Gmail for 3 billion people. The question is not whether your school has Google Cloud. The question is what you’re doing with it.

For most schools, the honest answer is: almost nothing.

Attendance gets entered into a proprietary gradebook that cannot export its own data. Student assessments live in spreadsheets scattered across 60 teachers’ Drives with no connection to each other. Automation means someone copy-pasting from one system to another. And every August, the school renews a contract for legacy software that does 10% of what it already pays Google to provide — because that software has a local helpdesk, a familiar interface, and a procurement officer who approved it three years ago.

This article is about the other 90%.


Layer 1 — What you already have: Google Workspace as a development platform

Most school administrators think of Google Workspace as a suite of apps. Technically, it is a managed cloud platform with a full-stack development environment built in. The distinction matters enormously.

Apps Script — your serverless backbone

Google Apps Script is not a “macro tool.” It is a V8-powered JavaScript runtime that executes server-side, with native bindings to every Google service your school uses: Drive, Sheets, Docs, Gmail, Calendar, Classroom, Forms, Groups, Admin SDK. It supports HTTP requests, scheduled triggers, event-driven webhooks, and service account authentication.

In practice, this means you can build automation pipelines that run without any infrastructure cost, on a schedule, in the background, connected to real data. We use Apps Script to run processes that — on any commercial SaaS platform — would require a monthly subscription, an integration consultant, and a three-month implementation timeline.

Examples from production at our school: automated newsletter delivery to 750 families with GDPR-compliant BCC routing; a 3D print queue system that receives, validates, and schedules student submissions; event registration pipelines that generate documents, send confirmations, and create calendar entries — all triggered by a single Google Form.

None of this requires a developer on staff. It requires someone who is willing to learn, and Google has already made that easier than almost anything else in the ed-tech market.

The Admin SDK and Directory API

Your Google Workspace admin console is a graphical interface over a fully documented REST API. Every user account, group, Chromebook, device policy, and organisational unit in your domain is queryable and modifiable programmatically. This means student onboarding, account lifecycle management, and device configuration can be automated from day one — not managed manually by an admin who spends two weeks every September clicking through a console.

“The question is not whether your school has Google Cloud. The question is what you’re doing with it.”


Layer 2 — One step further: Google Cloud Platform proper

Activating a Google Cloud project linked to your Workspace domain takes approximately ten minutes. What opens up is a different category of capability entirely.

Compute
Compute Engine (VM)

Full Linux or Windows virtual machines on Google’s infrastructure. Run any software, any daemon, any custom service — without owning hardware. Ideal for school automation servers, local AI endpoints, or any workload that needs always-on execution.

Containers
Cloud Run

Serverless container execution. Deploy any Dockerised application and pay only for actual invocations. Zero infrastructure management. Perfect for school-built microservices, APIs, and web applications that need to be public but don’t justify a full VM.

Data
BigQuery

Petabyte-scale analytics SQL engine. Free tier covers 10GB storage and 1TB of queries per month. Feed it attendance, assessments, behaviour logs, sensor data. Query it in seconds. The difference between a gradebook and institutional intelligence.

Visualisation
Looker Studio

Live dashboards connected directly to BigQuery, Sheets, or any Google data source. Not a static export. Not a PDF report from last Thursday. A real-time view of what is happening in your school, updated automatically.

AI / ML
Vertex AI & Gemini API

State-of-the-art foundation models accessible via REST API. Content moderation, automated summaries, smart classification, document generation — all callable from Apps Script or any backend you choose. An AI layer you control.

Storage
Cloud Storage

Object storage for files, backups, media, datasets, model artefacts. Globally distributed, versioned, fine-grained access control. Integrate with Apps Script to store generated documents outside the Drive quota, at cents per GB.

Database
Cloud SQL / AlloyDB

Fully managed PostgreSQL or MySQL. If your school builds any application that needs relational persistence — a custom registration system, a resource booking tool, a custom e-register — Cloud SQL removes the database administration burden entirely.

Integration
Pub/Sub & Cloud Functions

Event-driven messaging and micro-execution. Trigger a function when a file lands in Cloud Storage, when a student submits a form, when a sensor sends a reading. The glue layer for any real-time automation architecture.

Practical note: Every service above operates under Google’s EU data residency options and is covered by the same DPA your school signed with Google Workspace. You are not leaving your data sovereignty framework. You are extending it.


Layer 3 — AI, natively: Gemini in the school stack

The conversation about AI in education tends to orbit around one question: “What AI tool should we buy?” That is the wrong question. The right question is: “How do we integrate AI into the systems we already operate?”

Gemini is not a product bolted onto Google’s education offering. It is an API endpoint — generativelanguage.googleapis.com — callable from Apps Script, Cloud Functions, or any server you control. This means AI can live inside your processes, not as a separate tab students open to write essays for them.

Concrete implementations we run or have built: Gemini-powered content moderation on student 3D print submissions. Daily AI-generated commentary on air quality sensor data, formatted and pushed to a Looker Studio dashboard. Smart document generation from structured form data. A school-specific RAG assistant built on school documents, answering parent and student questions about procedures, timetables, and regulations — in Polish, with correct context.

None of these required a data science team. All of them required someone who understood that the API was there and decided to use it.


The infrastructure problem nobody talks about

Schools pay significant sums every year for software that cannot do what Google Cloud already provides — and cannot connect to it either. Legacy school management systems were built when “the cloud” meant “someone else’s server.” They store data in proprietary formats. They generate reports through a fixed interface. They cannot be automated, extended, or integrated.

Meanwhile, the school has a Google Workspace tenant with a documented REST API, an Apps Script environment, access to Gemini, a BigQuery sandbox, and a free-tier Cloud project waiting to be activated.

The gap is not money. It is not infrastructure. It is knowledge — specifically, the knowledge that this is possible and the institutional will to pursue it.

“The gap is not money. It is not infrastructure. It is knowledge — and the institutional will to pursue it.”

A school that treats its Workspace tenant as a platform — not a collection of apps — can build systems that are more capable, more integrated, more maintainable, and more aligned with actual pedagogical needs than anything a commercial vendor will sell it.


The skills pathway — from educator to cloud practitioner

Google has built a skills pathway that covers exactly this terrain — and it starts where teachers already are.

LevelCertificationPlatformWhat it covers
EducatorGoogle Certified Educator L1skillshop.withgoogle.comCore Workspace tools in teaching contexts. The entry point.
EducatorGoogle Certified Educator L2skillshop.withgoogle.comAdvanced Workspace workflows, integration, differentiated instruction.
EducatorGoogle Certified Trainerskillshop.withgoogle.comQualifying to deliver Google’s official educator training.
InnovationGoogle Certified Innovatoredu.google.com/innovatorsProject-based. Cohort model. Global network. Requires application.
CloudGoogle Cloud Digital Leadercloud.google.com/learnNon-technical cloud fluency. Essential for school leaders making platform decisions.
CloudAssociate Cloud Engineercloud.google.com/learnDeploying and managing GCP workloads. The technical foundation for everything in this article.
ProProfessional Cloud Architectcloud.google.com/learnDesigning scalable, secure GCP solutions.
ProProfessional Data Engineercloud.google.com/learnBigQuery, Dataflow, Pub/Sub, data pipelines.
ProProfessional ML Engineercloud.google.com/learnVertex AI, model training, MLOps.

The practical path: start with Educator L1/L2 if you haven’t already. Move to Cloud Digital Leader to build strategic fluency. Then Associate Cloud Engineer for hands-on technical capability.

All preparation materials are free on skills.google.com and cloudskillsboost.google.com — real cloud infrastructure, pre-provisioned, with guided labs. You can learn BigQuery on a real BigQuery instance, at no cost beyond your time.

For school leaders specifically: Cloud Digital Leader does not require you to write code. It requires you to understand what is possible and why it matters — well enough to make informed decisions about what your school builds, buys, or does without. In 2026, that is a core leadership competency.


What this looks like in practice

I run innovation and educational technology at a private school in Szczecin, Poland. We are a Google for Education Reference School — roughly 750 students, 60 teachers, four school divisions. We do not have a dedicated IT department. We have one person who decided to treat our Workspace tenant as a platform and learn what that actually meant.

In the last eighteen months we have built: a fully automated 3D print queue with Gemini-powered content moderation and MQTT-connected Bambu Lab printers; an environmental monitoring pipeline ingesting data from an Airly sensor into Sheets, through BigQuery, into a live Looker Studio dashboard with daily AI-generated commentary; complete document automation for school events — from a Google Form submission to a generated document, a calendar entry, and a confirmation email, with no human in the loop.

None of these projects required external vendors. None required budget beyond Google’s free tiers and our existing Workspace subscription. All of them are maintained by the same person who built them, because that person understands the full stack — from Apps Script to GCP to the Gemini API.

That is what treating a school as a platform looks like.


The opportunity is real. The infrastructure exists. The learning path is documented and largely free.

The only thing standing between your school and a fundamentally more capable, more data-informed, more automated operation is the decision to stop treating Google Workspace as a file cabinet with email attached — and start treating it as the cloud platform it actually is.

Every Google for Education school in the world has this. Very few are using it. That gap will not last forever.