Please describe your proposed solution.

This project aims to deliver a simple but powerful and flexible utility (both as an application and as a library with API) that allows users to compare skillsets to target job descriptions. This capability has many use cases, such as allowing employers to evaluate the suitability of job candidates, allowing jobseekers to identify appropriate target jobs, and many more.

The project is part of LearnerShape’s “open source learning infrastructure”, which allows users to build customized learning and workforce applications using open source components (see <https://www.learnershape.com/blog/using-ai-to-build-worlds-best-open-source-learning-infrastructure> for a further description of this approach). The overall idea of this approach is to allow learning applications to be built for specific use cases, rather than the usual approach in the edtech sector of building platforms that have complex (and often confusing) features that seek to satisfy a large number of use cases in a generic way. By contrast, we expect that the straightforward skills-to-jobs comparison capability that this project will deliver can be applied individually to a wide variety of focused and customized use cases.

Below we first explain the overall background, vision and progress of LearnerShape SkillsGraph and LearnerShape’s lsgraph library, and then describe what we will deliver through this proposal.

LearnerShape SkillsGraph and lsgraph – background and vision

LearnerShape ‘SkillsGraph’ will be a fully-functional open-source, standards-based Web 3 skills management solution. It is built on three core technical pillars:

• blockchain authentication with credential privacy
• open source code with standard APIs
• standards compliance.

The proof-of-concept version of SkillsGraph was developed with Catalyst F6 funding and is available to the public at <https://skillsgraph.learnershape.com/>. Full documentation is available at <https://learnershape.gitbook.io/learnershape-skillsgraph/>. The open source code is available at <https://github.com/LearnerShape/ls-auth-api> (which provides the blockchain and authentication functions used by SkillsGraph) and <https://github.com/LearnerShape/ls-auth-ui> (which provides a front-end application that delivers these functions to users).

LearnerShape's vision is that SkillsGraph will provide an open-source alternative to existing skills management solutions. For example, the leading current solution is LinkedIn, which is highly functional, but a proprietary walled garden that controls user data. By contrast, SkillsGraph is open source so that any entity that wishes to do so can operate a SkillsGraph instance, and all instances can access standards-based user data (to the extent users wish to make it available) and blockchain proofs of skills authentication.

Separately from SkillsGraph, LearnerShape has developed ‘lsgraph’, an open-source AI-based library for recommendation of learning content and jobs using skills (with any skills taxonomy). This library is available at <https://github.com/LearnerShape/lsgraph>, and descriptions of its capabilities are available on our blog at <https://www.learnershape.com/blog>. As a key element of the vision set out above, we intend to combine this AI-based recommendation technology with the authentication technology of SkillsGraph (see below).

Deliverables – Extension of LearnerShape SkillsGraph and evolution of existing LearnerShape open source libraries

This project is based upon the simple insight that the capabilities of SkillsGraph and lsgraph can be enhanced and combined to deliver a highly-useful workforce capability that leverages the power of Web3 and AI technologies to compare skillsets to job descriptions. Specifically, the project will deliver the capability for a user to:

• generate a verifiable presentation of their CV/skillset, in a format compliant with World Wide Web Consortium (W3C) specifications,
• compare it to an arbitrary job description, and
• generate a similarity score.

We expect that users will be able to apply such similarity scores to in a wide variety of contexts to assess suitability of individuals for jobs in a wide variety of contexts. To ensure that these functions can be used in a robust and disciplined way, this project will also deliver associated governance frameworks, as described below.

We will deliver the following specific capabilities In this project:

• Blockchain / self-sovereign identity (SSI) functions – The current, proof-of-concept version of SkillsGraph allows individual skills to be recorded and authenticated by third parties in the form of W3C-compliant verifiable credentials (VCs), and to display a set of such VCs (i.e. a skillset / CV) via a web interface. We will enhance these capabilities as follows, using the latest features of Atala PRISM (which has improved substantially since the original version of SkillsGraph was developed):
• allow a skillset / CV to be presented as a W3C-compliant verifiable presentation (VP) – as for all LearnerShape skils applications, skills can be flexibly presented without the need to use a specific skills taxonomy (see <https://www.learnershape.com/blog/going-beyond-skills-taxonomies-with-AI> for an explanation of this approach)
• enhance the SkillsGraph VC specification for presentation of individual skills (whether as a standalone VC or part of a VP) – at present, these are limited to the skill name, skill description, type of skill (from the options of ‘education’, ‘work’, or ‘general’) and date of authentication; while this current functionality is highly flexible and can support any skill, more specificity (details to be determined during the project) will be useful for many applications
• provide the ability for SkillsGraph VPs and VCs to be exported for interoperability with other SSI applications.
• AI-based skills-to-job comparision
• We will add functions to SkillsGraph and lsgraph to allow a SkillsGraph VP to be compared to an arbitrary job description (formats for providing job descriptions are likely to include .pdf, .txt and/or .docx), generating a similarity score.
• We will generate the similarity score using publicly-available AI / machine learning models, which are likely to include large language models (LLMs) and/or embeddings models. The specific models to be used will be selected as part of the project.
• Governance frameworks – We will create:
• a governance framework for the skills-to-job comparison features of SkillsGraph, focusing on how SkillsGraph applies SSI and AI technologies to user inputs in order to deliver reliable skills-to-job comparisons
• a template governance framework for applications that use the skills-to-job comparison features of our open source libraries lsgraph and ls-auth-api.
• Other SkillsGraph features – We will also enhance other features of SkillsGraph to support its evolution from a proof-of-concept to a fully functional application:
• enhanced user authentication – This is currently based on email address, and will be extended to additional communications networks (e.g. Discord, Telegram and/or WhatsApp).
• enhanced UI / UX
• graphical interface — currently a simple interface using hyperlinks links and text
• simple, powerful user experience integrating above new features.

These capabilities will be delivered as both (a) new functions of SkillsGraph and (b) new capabilities of our open source libraries lsgraph and ls-auth-api that can be used in other applications via API interfaces.

Who will benefit - and how?

The functions delivered by this project will provide a powerful tool for comparing skillsets to any job description. We are not aware of any application that delivers such functions in a comparable way, in particular one that:

• is based around portable W3C-compliant VCs and VPs that are authenticated on a layer 1 blockchain
• allows skills to be recorded entirely flexibly without using a defined skills taxonomy.

For individuals, this will allow them to automatically compare their skills to job descriptions to establish closeness of match, allowing them to identify potential new job roles - and skills gaps - more effectively.

For organisation users (including organisations in the Cardano ecosystem and others), the new version of SkillsGraph and associated open source libraries will provide:

• more efficient team hiring and management by matching skills to job roles
• the option to develop bespoke skills-matching applications using LearnerShape open source libraries.

Both individuals and organisations will benefit from these enhanced capabilities of SkillsGraph:

• skills recorded as portable W3C-compliant VCs and VPs, with interoperability through export (full VC/VP interoperability will be delivered in future versions of SkillsGraph)
• enhanced VC specification
• enhanced user authentication
• enhanced UI/UX.

These benefits should be important to Cardano because they provide the potential for a highly useful service that is based on Atala PRISM and the Cardano blockchain, and is useful far beyond the Cardano ecosystem (LearnerShape intends future versions of SkillsGraph to be interoperable with other blockchains, and possibly also non-blockchain authentication platforms). This builds on the successful delivery of SkillsGraph through Catalyst Fund 6, and pilots of the SkillsGraph technology through Fund 7 and Fund 8 projects. More generally, such an application supports the wider adoption of the Cardano ecosystem for SSI purposes. More information on benefits to Cardano is in the response to the next question.

How does your proposed solution address the challenge and what benefits will this bring to the Cardano ecosystem?

The proposal is squarely aimed at the challenge goal of delivering applications based upon Atala PRISM, as well as “solv[ing] the business coordination and stakeholder challenge in parallel with the technology implementation effort”. We are building on our existing Atala PRISM application delivered in F6, and adding a widely-applicable AI-based, skills-to-jobs comparison capability, as well as making other technology and usability upgrades.

The proposal also meets the general requirements of the challenge as follows:

• Having a great team, product roadmap, milestones, and execution plan – Please see details in other responses in this proposal.
• Complete the Atala PRISM Pioneer Program – Team leader Maury Shenk has completed the Atala PRISM Pioneer Program.
• Complete a Governance Framework workshop – We commit to do this as part of delivery of the proposal, and would focus our participation in the workshop on the governance frameworks described above.
• Establish or join a Governance Framework working group at Trust over IP, or another host organization – We commit to do this as part of the proposal via Trust over IP. We have participated in the W3C Verifiable Credentials working group, which addresses governance frameworks as part of its activities.
• Identify stakeholders including issuers, holders, verifiers, policy makers, and their business requirements – We have identified these stakeholders in developing the proof-of-concept version of SkillsGraph. Issuers are individuals who self-certify their own skills and the third parties who authenticate those skills. Holders are the individuals who use SkillsGraph to record and compare their skills. Verifiers are third parties who wish to confirm the skills of an individual. Because SkillsGraph is a private, commercial service, policy makers have a limited relevance to it. We will further explore these roles in developing the governance frameworks as part of this project.
• Establish a cadence of meetings and contribution and convene a home for parties to adopt and participate in the Ecosystem and Ecosystem Governance – We have been regular participants in Cardano Catalyst community activities, including those associated with Atala PRISM. We would accelerate that participation in connection with this project.
• Progress towards publishing a Governance Framework – As described above, we would publish a governance framework and a template governance framework as part of this proposal.
• Develop a Proof of Concept, Pilot, or commercial launch – As described in detail in this proposal, our focus is to deliver a working skills-to-jobs comparison application (and associated open source libraries), augmenting the proof-of-concept SkillsGraph application developed in Catalyst Fund 6.

Benefits for the Cardano ecosystem include:

• Uses Atala PRISM for a useful and promising application with wide applicability for learning and workforce applications, introducing new open-source.
• Provides API-enabled capabilities that can be leveraged by other projects in the Cardano ecosystem (and more broadly), with a template governance framework for such implementations.
• Supports the wider adoption of the Cardano ecosystem for SSI/authentication purposes and brings together stakeholders to accelerate adoption and product-market-fit.
• Shares the business- and delivery-focused approach that the Project Catalyst team and IOG have articulated in connection with F10 – LearnerShape is an established company run by highly-experienced entrepreneurs, and is committed to delivering commercially valuable outputs in this project.

How do you intend to measure the success of your project?

This proposal is for a development project, and we will measure progress through delivery of the targets and milestones set out under Feasibility above. Our monthly reports will include progress reports against these milestones – we are currently using this approach for reporting delivery of our F8 project (which has been delayed but continues to make progress). By the end of this project, we aim:

• to have developed and deployed a version of LearnerShape SkillsGraph with the new functions described above and associated governance frameworks; and
• to have a fully-developed plan and roadmap for widespread and open interoperability of SkillsGraph with other applications for skills management and authentication based upon W3C standards.

Generally, to track the success of LearnerShape SkillsGraph, we will monitor the following KPIs:

• usage of instances of SkillsGraph operated by LearnerShape
• number of users
• number of credentials created
• number of credential authentications
• number of skills-to-job comparisons
• open source usage (of our Github repos)
• lines of new code
• number of contributors
• number of known instances of SkillsGraph.

Please describe your plans to share the outputs and results of your project?

We will share the outputs and results of this project in at least the following ways:

• deploying a new version of LearnerShape SkillsGraph for public use, with the additional functions developed in this project, including the ability to make skills-to-jobs comparisons
• making the source code for our project available on Github in new versions of our lsgraph, ls-auth-api and ls-auth-ui repositories
• discussing the project in meetings of SSI community groups, including the W3C Verifiable Credentials working group and the Trust Over IP community
• disseminating information in the Cardano community, including via Town Hall, After Town Hall and community groups such as Gimbalabs and Lidonation
• disseminating information via other LearnerShape marketing channels, including our blog, newsletter, social media posts (especially on LinkedIn) and event participation (including by Maury Shenk as a speaker).

What is your capability to deliver your project with high levels of trust and accountability?

There are two main types of evidence of LearnerShape’s capability to deliver this project: (a) our experience as a company in delivering this type of technology and (b) our track record on Project Catalyst projects. The detailed capabilities of the team members is described below.

Company Experience – As described above, LearnerShape is focused on building “open source learning infrastructure”. Among other things, we have delivered:

• the open source libraries lsgraph, ls-auth-api and ls-auth-ui (links provided above)
• a demonstration platform that uses the capabilities of lsgraph to recommend content and create learning pathways from a large content library (demo can be requested via <https://www.learnershape.com/contact>)
• private projects relating to skills management capabilities for companies and universities.

Catalyst Track Record

• F6: Universal Skills Authentication (<https://cardano.ideascale.com/c/idea/58915>) – This project was successfully delivered, resulting in the current version of SkillsGraph.
• F6: Crypto Regulation Surveillance (<https://cardano.ideascale.com/c/idea/59250>) – This small project was successfully delivered, and has resulted in team leader Maury Shenk becoming a key member of the larger DLT360 team.
• F7: PACE: Skills credentials (<https://cardano.ideascale.com/c/idea/60571>) – This is a pilot of the SkilsGraph technology on which LearnerShape has supported PACE. They project remains pending with PACE. LearnerShape has received no Catalyst funding for our work on the project.
• F8: PeopleCert DevOps Pilot on Cardano (<https://cardano.ideascale.com/c/idea/61940>) – This is a larger pilot of the SkillsGraph technology with an external entity PeopleCert (possibly expanding to a second entity Northeastern University), delivered in collaboration with ProofSpace (<https://www.proofspace.id/>). Although LearnerShape and ProofSpace have hit our milestones, the project remains pending because of delays at PeopleCert. We will deliver this project, and continue to submit monthly reports on progress.
• F9: LearnerShape SkillsGraph v2 (<https://cardano.ideascale.com/c/idea/63967>) – This was one of the highest-rated proposals in Fund 9 (average score 4.93) but was not funded due to heavy down-voting (as some proposals experienced in Fund 9). We are hopeful that the present proposal, which draws on the F9 proposal but substantially refocuses it on the skills-to-jobs comparison approach described above (including adding AI capabilities), will be approved by voters.

We have also engaged actively in the Catalyst community since late 2021. Our team leader Maury Shenk has completed the Atala PRISM Pioneer Program, our head of data science Jonathan Street participated in the Atala PRISM 2.0 beta program, and LearnerShape has developed a strong relationship with the Atala PRISM team.

What are the main goals for the project and how will you validate if your approach is feasible?

The goals of the project are as follows:

1. Deliver the enhanced application functionality specified in the proposal, including the skills-to-jobs comparison capability and the other technology deliverables specified under ‘Please describe your proposed solution’ above – Metric: Achieving the technology milestones set out below.
2. Prepare the governance framework and template governance framework specified in the proposal – Metric: Delivering final versions of each framework.
3. Promoting widespread adoption of the enhanced version of LearnerShape SkillsGraph – Metrics: See KPIs under ‘How do you measure the success of your project?’ above.
4. Delivering the stakeholder and community engagement activities specified in the proposal, particularly under ‘How does your proposed solution address the challenge and what benefits will this bring to the Cardano ecosystem?’ above – Metric: Delivery of the engagement activities specified in that section.

Please provide a detailed breakdown of your project’s milestones and each of the main tasks or activities to reach the milestone plus the expected timeline for the delivery.

Delivery of this proposal will involve the following workstreams:

1. SSI functions
2. Enhanced verifiable credential (VC) specification meeting W3C standards
3. Verifiable presentation (VP) specification meeting W3C standards
4. Delivery of the above, including integration of latest version of Atala PRISM
5. Adding code for the above to ls-auth-api, and updating associated documentation
6. AI functions
7. Choosing optimal AI methods for jobs-to-skills comparison (including similarity score calculation), expected to involving use of large language models and/or embeddings models
8. Adding code for the above to lsgraph, and updating associated documentation
9. UI/UX – skills-to-jobs management
10. Simple and highly-usable interface elements for:
11. Selecting skills for comparison and inclusion in a VP
12. Uploading job descriptions
13. Generating and saving similarity scores
14. Exporting VPs and VCs
15. Adding code for the above to ls-auth-ui, and updating associated documentation
16. UI/UX – authentication and other
17. Simple and highly-usable interface elements for:
18. Implementing additional authentication mechanisms (e.g. Discord, Telegram and/or WhatsApp)
19. Graphical interface
20. Adding code for the above to ls-auth-ui, and updating associated documentation
21. Governance frameworks
22. Delivering the governance framework for LearnerShape SkillsGraph
23. Delivering the template governance framework for other applications implementing our skills-to-jobs comparison functions
24. Stakeholder outreach
25. Participation in governance workshop through Cardano, and third-party governance groups (e.g Trust Over IP, W3C)
26. Outreach to Cardano community and our broader networks to generate interest in the capabilities that we are developing, and assessing stakeholder needs

This is a fairly complex project, which we plan to deliver over a period of approximately 10 months. Faster delivery would be possible in principle, but likely less efficiently due to the need for greater parallel operations of workstreams that have dependencies. The main targets for each month are set out below for the workstreams listed above (with milestones identified in bold). We will develop more granular project plans as the project proceeds – the LearnerShape team has extensive experience with project management.

October 2023

• funding awarded
• SSI functions: begin design of VC and VP specifications
• AI functions: high-level planning of approaches to skills-to-jobs comparison
• UI/UX – skills-to-jobs management: high-level planning of interface features
• Governance frameworks: governance stakeholder group outreach
• Stakeholder outreach: regular outreach to stakeholders in Cardano community and elsewhere

November 2023

• SSI functions: complete design of VC and VP specifications (milestone)
• AI functions: design and initial testing of approaches to skills-to-jobs comparison
• UI/UX – skills-to-jobs management: continued planning of interface features
• UI/UX – authentication and other: high-level planning of interface features
• Governance frameworks: continue governance stakeholder group outreach
• Stakeholder outreach: regular outreach to stakeholders in Cardano community and elsewhere

December 2023

• SSI functions: begin integration of new VC and VP functionality for Atala PRISM
• AI functions: complete initial design of approach to skills-to-jobs comparison (milestone)
• UI/UX – skills-to-jobs management: agreed interface design (milestone)
• UI/UX – authentication and other: continued planning of interface features
• Governance frameworks: draft governance framework for SkillsGraph (milestone)
• Stakeholder outreach: regular outreach to stakeholders in Cardano community and elsewhere

January 2024

• SSI functions: complete integration of new VC and VP functionality for Atala PRISM (milestone)
• AI functions: begin integration of skills-to-jobs comparison functions
• UI/UX – skills-to-jobs management: begin build of skills-to-jobs interface
• UI/UX – authentication and other: agreed interface design (milestone)
• Governance frameworks: continue governance stakeholder group outreach
• Stakeholder outreach: regular outreach to stakeholders in Cardano community and elsewhere

February 2024

• AI functions: complete initial integration of skills-to-jobs comparison functions in lsgraph (milestone)
• UI/UX – skills-to-jobs management: continue build of skills-to-jobs interface
• UI/UX – authentication and other: begin build of other UI/UX elements
• Governance frameworks: draft template governance framework for applications using skills-to-jobs comparison (milestone)
• Stakeholder outreach: plan for promotion of new SkillsGraph application and associated open source libraries (milestone)

March 2024

• SSI functions: create draft new documentation for ls-auth-api (milestone)
• UI/UX – skills-to-jobs management: semi-final implementation of skills-to-jobs interface (milestone)
• UI/UX – authentication and other: continue build of other UI/UX elements
• Governance frameworks: continue governance stakeholder group outreach
• Stakeholder outreach: begin implementing promotion plan

April 2024

• AI functions: create draft new documentation for lsgraph (milestone)
• UI/UX – skills-to-jobs management: integration of SkillsGraph functions with SSI and AI functions
• UI/UX – authentication and other: semi-final implementation of other UI/UX elements (milestone)
• Governance frameworks: semi-final versions of governance framework and template governance framework (milestone)
• Stakeholder outreach: continue implementing promotion plan

May 2024

• UI/UX – skills-to-jobs management: complete integration of SkillsGraph functions with SSI and AI functions (milestone)
• UI/UX – authentication and other: complete integration of SkillsGraph UI/UX elements (milestone)
• Governance frameworks: continue governance stakeholder group outreach, seeking comments on proposed governance frameworks
• Stakeholder outreach: continue implementing promotion plan

June 2024 – final integration and launch of all components and documentation

• SSI functions: finalise new functionality and documentation for SkillsGraph and ls-auth-api (milestone)
• AI functions: finalise new functionality and documentation for SkillsGraph and lsgraph (milestone)
• UI/UX – skills-to-jobs management: finalise new functionality and documentation for SkillsGraph and ls-auth-ui (milestone)
• UI/UX – authentication and other: finalise new functionality and documentation for SkillsGraph and ls-auth-ui (milestone)
• Governance frameworks: finalize governance framework and template governance framework (milestone)
• Stakeholder outreach: launch new version of SkillsGraph and open source libraries (milestone)

July 2024 – spillover month to handle possible project delays

Please describe the deliverables, outputs and intended outcomes of each milestone.

Please see response to the previous question for details of deliverables and milestones. With respect to each milestone, we will provide documentation and/or computer code sufficient to establish the progress specified.

To track progress against milestones and ensure that delivery remains on track, we will:

• designate a member of the team responsible for project management
• use standard project planning methods for tracking progress – we use Trello as our primary project planning tool
• hold regular meetings of our full team to discuss progress, issues and resolutions.

Please provide a detailed budget breakdown of the proposed work and resources.

The budget breakdown below matches the six workstreams identified in the previous response, plus funding for cloud server capacity. Time costs assume ADA 30,000 per person-month including overhead (which is well below cost for developers and managers at the level of seniority of our team), except that server costs are estimated AWS costs.

SSI functions: ADA 43,500

• Design: 0.25 person-months (pm)
• Coding: 1.0 pm
• Documentation: 0.2 pm

AI functions: ADA 36,000

• Design: 0.25 pm
• Coding: 0.75 pm
• Documentation: 0.2 pm

UI/UX – skills-to-jobs management: ADA 36,000

• Design: 0.25 pm
• Coding: 0.75 pm
• Documentation: 0.2 pm

UI/UX – authentication and other: ADA 36,000

• Design: 0.25 pm
• Coding: 0.75 pm
• Documentation: 0.2 pm

Governance frameworks: ADA 21,000

• Drafting: 0.3 pm
• Stakeholder outreach: 0.4 pm

Stakeholder outreach: ADA 18,000

• General outreach: 0.3 pm
• Release planning: 0.3 pm

Project management: ADA 15,000 (0.5 pm)

Cloud server capacity (on AWS): ADA 2750 – assumes $80/month for Elastic Beanstalk (including EC2 instances and load balancing), plus Relational Database Service, for 10 months

Total budget: ADA 208,250

Who is in the project team and what are their roles?

Our team members are:

• Maury Shenk, Founder & CEO (<https://www.linkedin.com/in/mauryshenk/>) – Maury is responsible for overall project design and supervision. He is an experienced entrepreneur, investor and lawyer with more than 30 years experience. He has strong technical experience in computer programming, machine learning and blockchain.
• Dr Jonathan Street, Head of Data Science (<https://www.linkedin.com/in/jonathanstreet/>) – Jonathan is responsible for implementation of LearnerShape's back-end solutions and technical infrastructure. He is a senior data scientist with over 10 years experience. He programs in various languages, focusing on Python.
• Dr Sean Miller, Head of Web Development (<https://www.linkedin.com/in/seanmiller1066/>) – Sean is responsible for implementation of LearnerShape's front-end solutions. He is an experienced senior full-stack developer with over 20 years experience, with a focus on Ruby on Rails.
• James Tuke, Senior Project Administrator (<https://www.linkedin.com/in/jamestuke/>) - James has lead responsibility for project management and administration. He has over 20 years of board-level experience in digital communications and online platform start-ups, including in the professional services, education and music sectors. He is a recognised problem-solver with a good eye for detail.

How does the cost of the project represent value for money for the Cardano ecosystem?

As noted above, the budget for this proposal is based upon a blended rate of ADA 30,000 per person-month, which at current exchange rates (as of 16 July 2023) is equal to approximately £88,000 per year. Taking account of our highly-efficient overhead costs of 10% (20% is standard), this is equivalent to a salary of approximately £80,000 per year.

While there is wide variation in developer and executive salaries in the United Kingdom (where LearnerShape is based, and Maury, Jonathan and James live) and Canada (where Sean lives), this amount is far below the salaries that the team members could obtain in other work. Salaries in the UK for highly-experienced developers often exceed £100,000 (and can be significantly higher), and salaries in the UK for executives like Maury and James are significantly higher still.

We recognize that the amount requested is not low, but it is clearly value-for-money for our experienced team, is less than our team is paid for other work, and does not include any element of profit for LearnerShape.