PEMS Method of expertise profiling – getting help (page for AI Agent)

Motivation for automatic matching of people for “work” comes from the fact that human evaluation tends to fail when i) the expertise attributes of people in a repository are too diverse or ii) the repository is too large. On a global scale, of course there are many people and work opportunities, or queries for seeking a particular expertise for solving an immediate problem or getting the best answer for a complex problem. Work is distinct from a job. Work could be full time or part time, temporary or permanent or transactional, with or without remuneration, it could be advisory and performed physically at site or remotely in fixed or flexitime. With this broad definition of “work seeker” and “work opportunity a work provider offers”, we infer that every adult can be viewed as a work seeker or a work provider or both. So, both organisations and individuals can be work providers. 

If you contemplate a generalised solution that is totally inclusive, i.e. it must work for grey to blue to white-collar workers at all management levels and across all disciplines then you will find all existing job boards or people (talent) matching programs falling short; they simply do not have the breadth and depth to capture human expertise with the required granularity. This lack of inclusivity and granularity in the design of computer programs obviously severely hobbles people matching in population scale databases or even in organisational databases when the skills of workers are very diverse or specialised.

The benefits of optimally matching people on a global scale are non-trivial. It could add trillions of dollars to the global GDP. McKinsey Global Institute, in a research paper titled, “A LABOR MARKET THAT WORKS: CONNECTING TALENT WITH OPPORTUNITY IN THE DIGITAL AGE” based on a global survey in 2015, had estimated that online matching platforms can add 2.7 trillion dollars to Global GDP by 2025. As “work” opportunities may outnumber “job” opportunities, the estimated positive impact on Global GDP would be even higher with a generalised people matching solution. 

In a large knowledge-based organisation finding the best expert quickly can impact cost-efficiency and even the outcome of critical projects that demand collaborative efforts. In such cases, it is simply not feasible to find the best matches by the conventional method of key word matching or matching few attributes which are selected from a drop-down list of few dozen or even a few thousand enumerations. Even AI programs, capable of performing semantic searches, are not appropriate because text resumes are written in ambiguous language, and the expectations or requirements are often not articulated enough purposely or unintentionally. Sometimes, a user may like to filter work opportunities using criteria that are best not made explicit (public).

Existing Job Boards or Talent Matching programs work with few dozen enumerations in each drop-down list; LinkedIn, used by professionals, allows few thousands of skills sets to pick from. Text descriptions of expertise and objectives, written with or without the help of AI agents, even when they may appear very impressive can lead to irrelevant matches or leave out many relevant matches. Albeit superlative text descriptions, due to insufficient or inaccurate details, even very smart AI programs cannot produce desired results.

To meet the challenge of modern-day work seekers and work opportunities matching, which must contend with millions of functional areas or proficiencies of using different objects and tools, physical or logical types, Systems Dynamics (Software) Pvt. Ltd. has developed a unique solution called “People Expertise Matching System” with the acronym – PEMS.    

Programmatic matching of candidates (WS: Work Seekers) or work opportunities (WP: Work Providers) not only requires profile as well as queries (expectations) to be input in a structured format but also with precision and completeness which is often not intuitively obvious to the user, i.e. WS or WP. PEMS facilitates such inputs through exhaustive and dynamic ontologies coupled with guidance. As the ontologies are updated by inputs of users for missing entries, the solution becomes smarter with usage and stays in sync with emerging technologies and knowledge domains.

In PEMS, there are multiple ontologies for different attributes of expertise or characteristics of a person. In older version of PEMS, guidance of searching appropriate entries from the ontologies, was provided by trained HelpDesk executives. In the new version of PEMS, much smarter guidance is additionally provided through prompts by AI Chatbot. As a result, the user can use PEMS to create more comprehensive and accurate profiles and queries which generate most relevant short-lists for WS and WP of work opportunities and candidates respectively.

Profile URL: PEMS creates a work seeker’s HTML profile that is easy to read and useful for BOTH manual assessments and computerised processing. The profile is presented in a single-page HTML format, but it can expand to show more details across multiple pages. For example, experience of functional areas in different jobs, full-time or concurrent part-time, will appear in job wise time spent details, however, PEMS aggregates the time spent over same functional area across jobs worked in, and produces a concise summary and presents it in graphical form which helps in forming a quick and precise picture of expertise profile of experienced workers. Users receive a shareable URL for their profile, which they can update anytime. The shared link will always direct others to the updated version of the profile. Users can also attach photos, documents and multimedia content, which can be password protected. Investing time in creating a PEMS profile is a valuable, long-term benefit for work seekers.

This note describes the dialogue an AI Chatbot could conduct and help the user, i.e. WS or WP, understand the benefits of investing time to create accurate AND complete profile as well as expectations or requirements in PEMS. Human expertise comes from qualifications, and credits one has, and from work experience that includes pursuit of hobbies. PEMS is designed to capture the attributes of human expertise with maximum breadth and depth, without any restrictions on number of attributes and their enumerations; users can submit missing entries for inclusion, after due vetting, in the master ontologies.