Hack2Work CV builder

As part of team 3 (football), this app looks to iterate on the experience which currently exists for claimants who are asked to submit a CV to Universal Credit.

Making a CV is a key part of finding a job. Why not make it a part of the UC experience?

What are the key points this focuses on?

As a claimant, I want a CV to fulfil my UC tasks
1. I do not want to write and re-write the same thing constantly
2. I want a free way to create a professional CV
3. I want a single process to make a professional CV
4. I would like targeted feedback on areas that I am lacking
As a claimant, I want advice on what jobs to aim for
1. I do not want to care about job sectors
2. I want to know which job areas are available locally/nationally
3. I want to know which roles I am fit for right now
4. I want to know which roles I could be fit for in 6 months
As a work coach, I want to reduce barriers to entry for creating a CV
1. I want all of my claimants to be able to create a CV
2. I want claimants to know why they are creating a CV
3. I want claimants to know what is important when creating a CV
4. I want give claimants a direction for where to create their CV
5. I want to help claimants help themselves with looking for a role
As a work coach, I want to focus on the most important issues when contact hours are used
1. I want to not focus on tedious fixes and revisions
2. I want to provide the most valuable insights to a claimant
3. I want a claimant to know how to improve their CV before coming to me
As a work coach, I want to give clear and targeted feedback to claimants CVs
1. I want a claimant to know which areas need improvement
2. I want to know when a claimant has revised weaker areas

What does it look like now?

A work coach uses UC Full Service to give a claimant several tasks to do along their journey for looking for work. One of these tasks includes creating and maintaining the claimant's CV for use in further job searching.

As part of this process, the work coach uses UC to generate a 'to-do' item to the claimant for them to submit their current CV for review. This is in the form of a simple file upload that a claimant posts back to their work coach to await further review and comments from that work coach.

The work coach may only have a 30-minute appointment to go through this CV and discuss any changes and or improvements with the claimant. Given that a work coach has several claimants to their name, there is not a lot of space to stretch that time for CVs that need extra help.

In the current COVID climate, any feedback is given over the phone for most cases, which compounds misunderstandings and confusion in its own way. While this may change in the future as COVID restrictions are wound down, it still remains as a one-time conversation where items can be lost or forgotten before improvements are made.

Once the CV is created for that claimant, it is then used in further job applications as required, but there is a lack of targeted advice on tailoring a CV to where an applicant would like to aim for in the job market - something which is a result of a single file-upload field.

A work coach follows some general ABC's of job searching for a claimant:

A: Any job - Anything is better than nothing
B: Better job - A job that the claimant enjoys or has interest in is better for retention and further development
C: Career (?) - A career is where the claimant is satisfied with their job and has room for growth

While the any job aspect of this is important for moving a claimant out of long-term unemployment, the other area that is important is ensuring that they are able to remain in that job. This can be seen as the better job stage, but there are still very few resources to help a claimant decide what they'd like that better job to be.

From the point of view of a work coach, the first time they see a claimant's CV is after the initial draft. Because a CV (by its nature) is an abridged overview of a person, there is very little which tells a work coach why a given detail is important enough to be included. A CV may only state that a claimant left academia in 2020 with no further job experience, compared to the reality of there being no jobs to take.

What could it look like?

The option that this team is looking to explore examines the idea of integrating a CV builder into the UC journey.

It looks to iterate on the current process by looking to alleviate several pain points that have come about from both the COVID pandemic. The two key areas of the problem statement that this looks to focus on are as follows:

Optimise the limited time spent with work coaches for claimants
Improve the experience for job seekers who have recently joined the market

Using interest and experience-driven questions, a potential CV builder would look to nudge a claimant towards tailoring their CV towards fields that they have interest in, or that they would like to work towards.

Once a claimant recognises the areas they would like to work towards, the service can recommend targeted areas for improvement. These may include a lack of skills, or experience. In addition, the service would look to automate some issues away, such as spelling errors or jargon recognition.

Description

The project is written in Node.js, and uses the following frameworks and technologies:

Framework/Tech	Info	Purpose
Node.js	https://nodejs.org/	Asynchronous event-driven JavaScript runtime
Restify	http://restify.com/	Framework for building RESTful web services
Bunyan Logger	https://www.npmjs.com/package/bunyan	JSON logging library for node.js services
MongoDB	https://www.mongodb.com/	MongoDB is a document-based distributed database
Mongoose	https://mongoosejs.com/	MongoDB object modeling for Node.js
Nunjucks	https://www.npmjs.com/package/nunjucks	View templating engine for javascript

Test Frameworks:

Framework	Info	Purpose
Chai	https://www.chaijs.com/	BDD / TDD assertion library
Sinon	https://sinonjs.org/	Standalone test spies, stubs and mocks for JavaScript
Mock-require	https://www.npmjs.com/package/mock-require	Simple, intuitive mocking of Node.js modules.
Import-fresh	https://www.npmjs.com/package/import-fresh	Import a module while bypassing the cache
Zombie	http://zombie.js.org/	Full-stack, headless browser testing

Project Structure

`src/router`

Module	Purpose
`middleware`	'pre' and 'on' handler chain for request/response handling
`routes`	controllers to register url listeners and provide model and page routing

`src/ts`

Typescript Classes for the project.

Module	Purpose
`objects`	Typescript objects for the project
`renderer`	nunjucks renderer initialisation
`logging`	Logging engine initialisation
`session-management`	Session handling

`src/views`

Contains nunjucks templates for the project

Folder	Purpose
`layouts`	Root layout and extensions
`pages`	Nunjucks templates for each page in the site, based on a particular layout template

`test/spec`

Unit tests for the project.

Folder	Purpose
`app`	Javascript unit tests
`browser`	Browser-based user journey testing, using the Zombie headless browser
`ts`	Typescript unit tests

Local Configuration

To run the project locally, the following configuration will be required.

certificates

Generate a self-signed certificate using npm run regenerate-certs, which creates the following two files:

certs/localhost-cert.pem
certs/localhost-privkey.pem

local.js

Create a new file config/local.js with the following contents:

  app: {
    http2: {
      enabled: true,
      key: 'certs/localhost-privkey.pem',
      cert: 'certs/localhost-cert.pem',
    },
    name: 'template',
    hostname: 'localhost',
    port: 8000,
  },
  nunjucks: {
    options: {},
  },
  logger: {
    level: 'info',
  },
};

Docker Configuration

The project contains Docker configuration files that allows the service to be easily and quickly started in Docker containers.

The site Dockerfile - service.Dockerfile - will use npm to build and package the Node project to run the website in a Docker container.

There is also a 'docker-compose.yml' which should be used to coordinate building the 2 docker containers which comprise the local project infrastructure.

Issue the command:

docker-compose up -d --build

to build the following containers:

Container	Details
`service`	The Template website, accessible via `https://localhost:8000`
`mongodb`	MongoDB instance, seeded with some basic test data, mocked as a replica set

Mongo Transactions

Before the template site can be accessed locally, the MongoDB instance needs to be running as a Replica set (albeit with a size of 1). This is to enable transactions within the application, to ensure atomic persisting of changes across the Mongo collections.

This is handled by the additional container rs-starter, which should be started alongside mongodb and will poll the mongodb service until it has started. At which point, it will enable the replica set of the database and close down.

Accessing the Website

The template can be accessed locally via: https://localhost:8000

J4Numbers / hack2work-cv-builder