Machine learning and blockchain-backed optimized assignment matching for PSWs to improve understaffing and risk during the COVID-19 outbreak

新冠肺炎疫情暴露了加拿大个人支持工作者(PSW)行业运营的系统性弱点。缺乏有效的认证证明和微凭证(急救、安全培训、背景调查等)导致招聘延误、人手不足，并给患者带来潜在风险。这反过来又给PSW带来了一些问题，包括低工资、低效率的日程安排、就业不足，并导致了不可持续的PSW“零工经济”。TriNetra正在与ConnexHealth合作实施一个系统，以验证PSW的资格、成就和认证。他们提议与Seneca合作，扩展该系统的功能，以包括根据候选人的整个概况(包括证书、培训、地理位置、工作历史和可用性)将候选人与工作分配相匹配的功能。这将有助于减少护理机构由于难以验证潜在工人的资质而导致的人员不足，消除PSWs与其指定任务之间的旅行和技能不匹配，并通过更好的任务匹配、对培训的认可和减少旅行来提高PSW的生活质量，从而帮助解决COVID危机。 该项目最初将创建一个功能齐全的PSW社区门户网站，具有基于区块链的认证和基于机器学习/人工智能的匹配系统，以满足当前对PSW更好和更可靠的认证和安置软件的需求。该门户将自动创建单独的PSW分配和时间表，以涵盖其培训、可用性和地理位置的所有方面。该项目的头两个阶段将在项目的头12周内部署一份需求报告和优化的匹配系统，并在24周的项目完成时部署一个功能齐全的系统。这将提供立竿见影的好处，帮助减少长期护理机构人手不足的情况，并帮助社会工作者在更匹配的任务中更快地就业。

The COVID-19 pandemic exposes systemic weakness in how the Personal Support Worker (PSW) industry is operated here in Canada. Lack of validated proof of certifications and microcredentials (first aid, safety training, background checks, etc.) leads to hiring delays, understaffing, and potential risks to patients. This in turn causes issues for PSWs, including low pay, inefficient scheduling, underemployment, and has resulted in an unsustainable PSW "gig economy." TriNetra is collaborating with ConnexHealth to implement a system for verifying PSW qualifications, achievements, and certifications. They are proposing to collaborate with Seneca to extend the features of this system to include capabilities to match candidates to job assignments based on their entire profiles, including certifications, training, geography, work history, and availability. This will help the COVID crisis by reducing understaffing at care facilities due to difficulty in validating the credentials of potential workers, eliminating travel and skills mismatches between PSWs and their given assignments; and improving PSW quality of life through better assignment matching, recognition for training, and less travel. This project will initially create a full-featured PSW community portal with blockchain-based credentialing and a machine learning/artificial intelligence-based matching system to address the current need for better and more reliable credentialing and placement software for PSWs. The portal will automatically create individual PSW assignments and schedules to encompass all aspects of their training, availability, and geography. The first two phases of the project will result in a requirements report and optimized matching system to be deployed within the first 12 weeks of the project, and a fully functional system at completion of the 24 week project. This would provide immediate benefit to help reduce understaffing in long-term care facilities and assist PSWs in faster employment at better-matched assignments.