Xopolis: Beyond Resumes

A Bayesian Deep Learned skills-to-task ontology for job recommendation and career pathing.

Worker resumes and employer job posts have been the standards for matching labor supply with demand at least since the Industrial Age, predating the computational advances of present day. These formats are structured to be concise, ideally with a high degree of specificity for the work they represent as they were designed for human readers. Yet, these formats have also become the primary input for job recommender systems. As a result, job matching based on computer algorithms continues to under-perform the manual approaches actually practiced by recruiters and hiring managers today, viz., Boolean searches on online candidate professional and social profiles coupled with an over-reliance on referral networks. The inefficiencies of this system are apparent in the increasing time-to-hire across most segments of the job market that is costing employers thousands of dollars a week in unfilled positions, and the fact that educators and policy-makers continue to report a lack of available data for precisely measuring the 'skills-gap' for developing retraining and upskilling curriculum in a timely manner.   

Xopolis’ solution is to introduce a high-resolution skills-to-tasks ontology that relates the skills that workers actually possess to the workplace tasks employers need performed. Such an ontology, complemented with a technological infrastructure for the aggregation and transformation of jobs data in near real-time, would track the evolution of work, provide invaluable insight on the future of work, and open up a new field of job-informatics.

Xopolis' solution develops a technology for analyzing and processing data on the work-content of jobs. The proposed platform aggregates data on dimensions of work not available in resumes or job posts by sourcing granular information on work content directly from recruiters and job seekers. Big Data techniques will be applied to this data to inform a jobs ontology that at its basis relates work tasks to worker skills. This new ontology along with proprietary algorithms underlying Xopolis’ platform will address foundational measurement issues and improve the construct validity of input data for job recommending, search, and matching algorithms.

This project will establish new technical processes for the sourcing, processing, and analysis of Big Data for job-informatics. Xopolis’ database will track close to 74,000 jobs that vary by occupation, industry, or geography along nearly 400 different dimensions. Data at this level of resolution, updated in real-time, possess a risk for computational systems which have so far not been applied to jobs data at the scale proposed. Furthermore, Big Data analytics will play an integral role in developing the skills-to-tasks jobs ontology underlying the platform. The success of this effort would prove a technological disruption to the study of work. 

Over the next 6 months Xopolis’ goal is to open access to the personal career profile generator, and the career mapping service for professional job seekers in the Washington D.C. metropolitan area. Access to the recruiter portal will also become available at this time but will be limited to partner recruiting agencies only, of which currently 1 team is actively engaged. Between 6 to 12 months, the company’s goal is to pilot and improve on the quality of user experience while onboarding over 15,000 passive candidates through the career mapping service and the personal career profiler. 

The company’s goals upon successful completion of the pilot are three-fold; 1) open the current services to US cities outside of Washington DC; 2) engage with job centers and local workforce boards to develop services for non-professional occupations; and, 3) work with institutional partners to develop industry standards for personal career profiles, as a replacement for resumes. Xopolis’s vision is to become a full-service platform integrating solutions for job seekers, employers, educators, and workforce policy makers. To do so, Xopolis will collaborate with a wide range of institutional partners to incorporate services for skills assessments, credentialing solutions, and social networking.  

To reach professional job candidates, BeyondResumes will market its online Career Mapping tool to career counseling services and professional associations in the Washington DC area. The same service will be offered for free to partnering recruiting agencies to extend to their network of candidates and job seekers. To reach and retain recruiting agencies, Xopolis is offering an initial partnership with a shared intermediation fee, based on the value proposition that our technology can halve the time it currently takes them to find and onboard passive candidates.   

Xopolis is currently in early stages of developing a pilot for a professional talent recruiting solution in the Washington DC metropolitan area that was identified during our customer discovery interviews. This project involves a small recruiting agency experienced in placing legal and professional talent in mid to top-tier positions and the pilot is being designed for a local IT consulting company that outlined its specific hiring needs over the next 6 months. 

In 12 months, Xopolis’ goal is to have at least 5 recruiting agencies actively using the BeyondResumes platform, each placing 10 candidates a month through the platform. We plan to achieve this scale by continuously improving on the Career Mapping tool and improving our algorithms according to the feedback we receive from recruiters and job seekers in our first year of growth. In 3 years, we anticipate out services to be available online and across most occupational categories and industries in the country. The success of Xopolis’ solution will be reflected in shorter time-to-hire for users of the platform.

Dani has technical capabilities for leading design and development of this solution. He has led the successful completion of substantial projects under grants from the Ewing Marion Kauffman Foundation and the National Science Foundation. In addition, Dani successfully completed the National Science Foundation’s national I-Corps program as the Entrepreneurial Lead and is versed in customer discovery. 

Auerswald has entrepreneurial experience having founded and co-founded four companies that span publishing, public policy, and land administration. Auerswald also provides the business development leadership for connecting Xopolis with entrepreneurial resources, and the strategic planning for developing the necessary ecosystem around the BeyondResumes solution.

Xopolis’ business model is to collect an intermediation fee for every successful hire made through the platform. The platform will offer free services for job seekers and career changers to translate their resumes and work histories into personalized career profiles designed for advanced career exploration, and in the process onboard talent onto the platform. Recruiters and hiring managers will use Xopolis’ job recommender system to produce lists of talent that is matched and ranked to employer needs. The platform will encourage recruiters to build employer profiles and anticipate hiring needs over a 6 to 12-month period. In return, they also receive lists of candidates that do not currently have the requisite skills but can be upskilled with the required time. The identity of the talent is maintained as pseudonyms until both the candidate and the recruiter verify interest in an interview. Upon a successful hire, Xopolis will collect an intermediation fee of 1 percent of annual salary of the position. In addition, the company will also develop data products based on the plethora of job market information it will be collecting from its users. 

Xopolis’ solution to democratizing opportunity in the job market is to fundamentally change the current standards on how job search and match are performed. This proposed disruption is achieved by replacing resumes with personal career profiles for job candidates and replacing job posts with project specific employer-needs profiles. Such an update would also collect previously unavailable data to study work. To successfully achieve disruption, Xopolis needs to collaborate with a range of institutional partners in industry, education, and government. An affiliation with Solve will help open doors to partners and collaborators that can help map, design, and accomplish Xopolis’ goals.  

The technical risk inherent in attempting a solution at this scale is a key challenge. BeyongResumes will track over 74,000 jobs varying in specificity by occupation, industry, and geography along nearly 400 work descriptors, 19,000 initial task categories, and 34,000 individual skill descriptors. Our algorithms were developed and tested on a subset of the data, but analysis at this scale possess a risk for computational systems which have not yet demonstrated viability at the scale proposed. Solve can help connect us with individuals experienced in developing information technologies at this scale and to assess the feasibility of this venture.