HR Technologies

HR Technologies

Wistron began its digital transformation in 2014 and has been cultivating digital talents systematically and applying digital technologies in various business fields. HR technologies refer to the utilization of digital tools to better automate talent selection, hiring, cultivation, and retention. Data analysis and AI technology can help the HR department better understand the company's human capital, predict and analyze data, which can help the HR department better manage decision insights and trends and thus identify the company's risks and opportunities more effectively. In the future, Wistron will devote more resources to building a comprehensive talent development and skill management system. By tracking employee skills, planning employee individual development, integrating and recommending courses, the company can better identify employees with great potential, develop excellent leaders, and continue to drive employee growth with the help of data analysis and machine learning.

Digitalize Recruitment

In this competitive talent market, it is crucial to hire the right talent and do so fast. Wistron has utilized technologies to simplify and integrate the complicated recruitment process that spanned across multiple systems, and build a one-stop recruitment system. This recruitment system officially began operation at the end of 2022 in Taiwan. The platform not only includes internal/external talent pools, but it also helps HR supervisors stay on top of the entire recruitment process, reducing the talent selection time by approximately 41%. In the future, Wistron will continue to optimize the system based on the company's needs and gradually adopt it at our overseas locations to complete our digitalized recruitment plan worldwide.

 
 

The recruitment system integrates 3 major AI models and generative AI applications to help supervisors find the best talent
	WiCruiting resume matching	NLP behavioral competence assessment	HRDA digital competence assessment	Smart JD content generator
Targets	Expedite supervisor's resume processing speed and reduce the time and workload of manual resume filtering.	NLP (Natural Language Processing) language analysis technology can analyze an applicant's behavioral competence, providing supervisors with a reference that can aid in making hiring decisions, which helps them save time and reduce recruitment costs.	AI facial recognition and micro expression analysis, combined with machine learning, can analyze an applicant's six major digitalized development potential indicators for supervisor's reference before the interview.	By combining generative AI (Gen AI) with new hire applications, supervisors can speed up writing job descriptions (JD) with greater accuracy.
Method	Conduct a correlation analysis between the resumes of newly hired employees and the company's job openings and build an AI resume-matching model (WiCruiting). To allow supervisors to view all resumes on the same interface, Wistron has incorporated the resume recommendations from external job websites into our recruitment system this year and connected the system to WiCruiting to generate resume-matching scores. Resumes with high matching scores will be recommended to the supervisors of corresponding departments or the HR staff, enabling them to quickly identify the most suitable candidates.	NLP emotion analysis, coupled with voice recognition technology, can quantify more than 100 voice features (such as tone, volume, long/short sentence ratio) during actual interviews between applicants and supervisors. These voice features, along with the assessment results of behavioral competence, are then be used to develop an AI model.	The digital interview (HRDA) system had captured the facial expressions of Wistron workers as they responded to questions, employing AI facial recognition, micro-expression analysis, and machine learning to develop a dedicated digital competence evaluation model for Wistron. We've also integrated video interviews into our recruitment process, allowing us to capture micro-expressions from 86 points on job applicants' faces. These micro-expressions will be analyzed alongside competencies to predict candidates' suitability, providing supervisors with objective references for making hiring decisions	By leveraging agile development practices, we rapidly integrated generative AI (Gen AI) into our recruitment system. Utilizing ChatGPT, we gathered job description data from both internal and external sources within Wistron. This approach allowed us to generate clear and comprehensive job descriptions and requirements for positions proposed by supervisors.
Results	Compared to the resume matching function developed jointly by Wistron and external job banks, WiCruiting demonstrates a 4-fold increase in efficiency in identifying a suitable resume.	Wistron has utilized this technology in our selection of reserve associates as well as our recruitment interviews. Via NLP, we analyze the speech interaction between applicants and supervisors to effectively predict the 6 major behavioral competence indicators of each applicant and quantify 9 speech features that serve as references for supervisors during the hiring decision-making process. It has been implemented to each year's reserve associate selection starting in 2021 with a total of 305 applicants assessed so far.	Since its introduction in 2021, Wistron has utilized the HRDA system to analyze the digital competencies of about 8,900 applicants. With continuous iteration of the system and the incorporation of speech analysis, the model now has approximately 69% in its predictive ability with about 96% of our supervisors considering digital competence assessment valuable when making hiring decisions.	Improved supervisor's writing efficiency by 50% and reduced the back-and-forth between the HR and supervisors regarding a position's duties and requirements by 75%.

Connected and Automated Data (RPA)

Integrating the talent information both inside and outside the company and connecting with other platforms' data to ensure data completeness and consistency. Digital tools and systems make HR process automation a reality.
 

Targets	Method	Results
Enhancing HR's internal operations by automating and digitalizing repetitive HR tasks. This approach not only improves work efficiency but also ensures data accuracy across various HR processes.	Incorporate Robotic Process Automation (RPA) technologies and conduct a thorough review of the current HR processes to formulate and then implement a more comprehensive, simple and standardized operational process. We had developed multiple RPA tools specifically designed to automate regular and repetitive manual operations, significantly reducing time spent on such activities.	By standardizing the execution of similar HR processes, Wistron can reduce work hours on repetitive work and increase HR's production value.

Data Visualization

 

Data Modeling, Prediction & Analysis (AI)

In addition to the 3 major AI models - WiCruiting resume matching (our 2.0 recruitment system), NLP behavioral competence assessment and HRDA digital competence assessment, Wistron has developed a resignation risk prediction model to address employee retention issues. This model aims to identify early signs of an employee's inclination to resign, thereby enhancing the company's ability to retain talent and mitigate the risk of talent loss.

Prediction of employee resignation risks

For indirect employees, Wistron has started actions to use data technology to improve the turnover rate. Relevant projects and results are summarized in the table below.
 

Featured project	Risk Prediction of Employee Resignation
Purpose of the project	The employee's inclination to resign is predicted through the integration and analysis of the company's internal and external information; therefore, active retaining actions can be taken in advance.
Project benefits	The company predicts the likelihood of key talents resigning, enabling proactive measures to be taken for their retention. What were once passive responses are now transformed into active strategies, aimed at enhancing overall talent retention within the company.
Data collection	70 personnel-related data entries from within the company.
Project results	Since the introduction of the project in Taiwan in 2018 Q4, around 1,300 high-risk personnel have been reported, with an overall prediction rate of above 70%. As a result, the indirect employees' voluntary turnover rate in the Taiwan headquarters has been decreasing year over year, dropping from 13% in 2020 to 10.55% in 2023.