Jump to Content
Developers & Practitioners

How Rhode Island created a virtual career center

March 12, 2022
https://storage.googleapis.com/gweb-cloudblog-publish/images/VCC_Thumb.max-1300x1300.jpeg
Alexandrina Garcia-Verdin

Google Developer Advocate

2020 brought many challenges, especially as in-person operations were shut down, and many were left vulnerable to unemployment.

The State of Rhode Island responded to these challenges, by modernizing their workforce development operations and moving completely to a custom online platform called the Virtual Career Center, also  nicknamed “the VCC.”

https://storage.googleapis.com/gweb-cloudblog-publish/images/VCC.max-1800x1800.png
Snapshot of the Virtual Career Center

(Click to enlarge)

It was developed in partnership with Google Cloud and a nonprofit called “Research Improving People’s Lives” (RIPL), and was fully built by Google Cloud partner Maven Wave, which helps a wide array of organizations, including public sector customers, with many types of cloud initiatives.

https://storage.googleapis.com/gweb-cloudblog-publish/images/MavenWave.max-1500x1500.png
List of Maven Wave’s services

(Click to enlarge)

In this episode of Architecting with Google Cloud, we interviewed Joel Osman, head of Digital Experience & Custom Applications at MavenWave who shared a lot of insights such as:

We looked at how we can apply leading edge emerging technologies to help people get back to work and use them as tools in such a way that we can augment the personal one to one interactions that agencies had been using with job seekers, to help them get back to work.

One of those key benefits is enabling job seekers to find coaches that are specialized in their respective community. During the in-person walk-in model, applicants were paired on a first come first serve basis with any available coach on site. Meanwhile online scheduling has enabled a greater opportunity to match veterans, college graduates, non-English speakers, etc with coaches with prior experience in that specific area.

How the VCC was built

This VCC web app was built on Angular. It has a custom frontend built on top of 2 key Google Cloud products. The first is Workspace, which includes functionality such as video conferencing, documents, slides, chat, file storage, etc. And the other is Google Cloud computing resources. Here’s a view of the architecture:

https://storage.googleapis.com/gweb-cloudblog-publish/images/VCCArch.max-1000x1000.png
Architecture of Job Coaches, Job Seekers, and Employers interacting with Google Cloud and Workspace architecture.

 There are 3 main types of users at this time, and that’s job coaches, job seekers, and employers.

  • ?‍? Job Coaches all have Google IDs in the Google Workspace domain, and therefore authenticate against the Google identity repository.

  • ?️‍♂️ Job Seekers are authenticated through a Cognito-based process maintained by the nonprofit I mentioned earlier (RIPL), the Rhode Island infrastructure team, and the Department of Information Technology (DoIT).  Cognito was an identity repository setup prior to this project for users interacting with the State, and remained as their form of authentication.

  • ? Employers participate directly with Google Meet, and, to an extent, Google Calendar; but not the Angular app. There’s also a focus on building a future dashboard to see how the center has helped employers with applicants.

The specific Google Cloud components used are the following:

  • Firestore: realtime Database that keeps data in sync across client apps.

  • BigQuery: serverless warehouse for data.

  • Data Studio: is used to build filterable dashboards over BigQuery

  • Cloud Functions: which serve as triggers to keep scheduling and data workflows in sync.

  • Kubernetes cluster: runs & autoscales the server-side code in a single-region deployment, with a minimum of four nodes per zone across three zones of the US East region.

  • Cloud Armor: protects applications and websites from attacks, and sets NIST-compliant policies.

  • Google’s Content Distribution Network (CDN): content is accessed and cached.

To manage the lifecycle of the infrastructure, a Terraform script is used, which is an open source  tool, and is structured into 5 folder environments: 

  • Admin

  • Dev

  • QA & UAT

  • Networks

  • Prod

  • Shared Services (for CI/CD pipelines between Dev & Prod).

Adoption outcomes

A universal fear we technical practitioners may have is:

Will our tool be loved and adopted by our intended audiences?

 Joel mentioned Job Coaches at the time were not used to working from home, and the team was concerned that they would potentially feel overwhelmed with a lot of new technology.

 As a rewarding surprise, when Job Coaches were presented the proof of concept, it was received with positivity.

95% of Job Coaches rated the VCC as a valuable solution and 87% reported to find it very or extremely effective.

This alignment was thanks to designing the tool with the users in mind, and performing user research since the beginning of the journey, which helped address their day to day needs.

https://storage.googleapis.com/gweb-cloudblog-publish/images/Rhode_Island_Job_Coach_Page.max-1900x1900.png
Screenshots of Job Coach (left) and Job Seeker pages (right). The Job Coach page includes calendaring, resources, and shortcuts to take quick actions, The Job Seeker page includes upcoming meetings, past meetings, and job search history. Source: Maven Wave.

Additional innovation for the future

After creating a centralized hub and moving operations to a digital format, many more benefits also arise. For example, there can now be an integrated data analytics view which enables meaningful dashboards that can be customized for different audiences such as job applicants, coaches, program stakeholders, or state agencies. 

https://storage.googleapis.com/gweb-cloudblog-publish/images/Analytics_Dashboard.max-1500x1500.png
Screenshot of analytics dashboards by Maven Wave

There can be improved job searchability by integrating machine learning, helping with resume building and parsing that take keywords out of a resume and match them to a variety of relevant job clusters, rather than just performing raw keyword searches.

https://storage.googleapis.com/gweb-cloudblog-publish/images/Web_Mobile.max-1600x1600.png
Screenshot of ML natural language-based job searching by  Maven Wave

Embedding chat bots can also help reduce the load of call centers in states, as they utilize natural language processing as well to help guide job seekers with prompt answers. 

https://storage.googleapis.com/gweb-cloudblog-publish/original_images/Minessota.gif

Conclusion

The State of Rhode Island’s Virtual Career Center is an amazing success story. By having worked with an experienced partner to move their operations to a digital format, they were able to respond to their citizen’s needs in a time where in-person operations were not possible. They also unlocked opportunities such as better matching and reporting along that journey. 

For any organization whether they are in the public sector, university, private sector, etc; anyone can take advantage of this platform and customize it to their needs as Maven Wave shared that they offer a menu of options, where you can pick and choose functionality based on your requirements, IT resources, and budget.

Posted in