This repo contains participant information for the Telstra Innovation Day event. Use information in this repo to explore data and develop operational apps.
The Telstra GCP Innovation Day is an opportunity for Telstra staff to create and build and idea to solve business problems. The platform we will use is Google Cloud Platform which has foundations within Telstra.
There are two Streams of focus for the Innovation Day:
-
Application Modernisation - Kubernetes and container-based application modernisation. This will solve the time-to-market and cost problem of developing and bringing solutions to customers.
-
Data and ML Operations - Batch, Stream data processing architectures which implement the latest technology in model development, (re)-training and serving. This stream focuses on bringing actionable insights and real-time Machine Learning to big and fast data.
- Storage:
gs://telstra-id-PROJECT_ID-storage/ - Git:
gcloud source repos clone --project=telstra-id-PROJECT_ID - GKE creds:
gcloud container clusters get-credentials --project=telstra-id-PROJECT_ID --zone="australia-southeast1-a" "telstra-id-PROJECT_ID-gke"
There are five data synthesizers which are streaming data for available for use. The synthesizers are each publishing into their own Pub/Sub topic. The messages can be retrieved from the associated Pub/Sub subscription.
These streams are available:
There are two generators that run for CDRs. The Australian CDR generator will
create call records to and from ten digit numbers from Australian area codes
and mobiles. Published to PubSub topic cdr_aust. The International CDR
generator will create call records from Country Code + Nine digits numbers, the
CDR will always have and Australian (+61) number as either the call origin or
destination. Published to PubSub topic cdr_intl.
This synthesizer attempts to simulate alerts that are typically generated by
datacenter servers and networking devices. Alerts are published to the topic
alarm_synth and are available in the subscription alarm_synth.
This generator will simulate cell tower data in JSON format. Data is published
to a topic called celltower_synth, there is an associated subscription called
celltower_synth.
Customer modem health responses are simulated with the data schema below.
| Column name | Description |
|---|---|
| customer_id | ID of the customer whose modem is online |
| timestamp | Timestamp of when modem ping returns |
| suburb_state | Suburb and State for the address of the customers' premise |
Service requests for impacted services are simulated using the data schema below.
| Column name | Description |
|---|---|
| customer_id | ID of the customer who submitted a service request related to a fault |
| timestamp | Timestamp of when the customer was impacted by their internet service being down |
| suburb_state | Suburb and State for the address of the customers' premise |
These datasets are available to teams and loaded either into your BigQuery instance, or available in a storage bucket.
- Storage bucket:
gs://telstra-id-storage - Storage bucket:
https://source.cloud.google.com/telstra-id/shared-resources
OpenCelliD provides a huge cell tower dataset used for a multitude of commercial/private purposes by patrons worldwide. Use this dataset for locating devices or understanding network coverage patterns. A worldwide cell tower database is available here:
We will also be providing data from Opensea specifically from the “Art Blocks Factory” project and making it available in BigQuery. This will include details for the NFTs available as well as historical sales data. For info regarding the project as well as the data schema see:
| Column name | Description |
|---|---|
| token_id | The token ID of the ERC721 asset |
| image_url | An image for the item |
| background_color | The background color to be displayed with the item |
| name | Name of the item |
| external_link | External link to the original website for the item |
| asset_contract | Dictionary of data on the contract itself (see asset contract section) |
| owner | Dictionary of data on the owner (see account section) |
| traits | A list of traits associated with the item (see traits section) |
| last_sale | When this item was last sold (null if there was no last sale) |
In addition to the supplied data sets, Google also have a repository of samples datasets which can be found here; https://cloud.google.com/bigquery/public-data
The data itself is located here; https://console.cloud.google.com/bigquery?project=bigquery-public-data&page=project
The dataset of most interest are:
- catalonian_mobile_coverage
- catalonian_mobile_coverage_eu
There are additional datasets provided in a bucket called telstra-id-datasets
within the shared project.
These are available as a simple one-click deployment for hosting applications on managed Cloud Run.
| Framework | Description | Deploy |
|---|---|---|
| React.js | React Sample | |
| Sapper.js | Sapper Sample | |
| Svelte Kit | Sveltekit with TailwindCSS | |
| Nuxt.js | Nuxt.js with TailwindCSS and TypeScript | |
| Next.js | Next.js with TailwindCSS |
Additional information and resources are available at the links below:
Telstra provides thier own github resource. Which contains many interesting APIs and code examples. https://github.com/telstra/
Event Detection API is a good example, where and API can be access on Enterprise Mobile phone accounts to access phone events. https://github.com/telstra/Event-Detection-API