Employee experience – the missing link for engaging employees: Insights from an MNE's AI-based HR ecosystem

Analyzing multiple data sources from a global information technology (IT) consulting multinational enterprise (MNE), this research unpacks the configuration of a digita-lized HR ecosystem of artificial intelligence(AI)-assisted human resource management (HRM) applications and HR platforms. This study develops a novel theoretical framework mapping the nature and purpose of a digitalized AI-assisted HR ecosystem for delivering exceptional employee experience (EX), an antecedent to employee engagement (EE). Employing the theoretical lenses of EX, EE, AI-mediated social exchange, and engagement platforms, this study's overarching aim of this article is to establish how AI-assisted HRM fits into an organization's ecosystem and, second, how it impacts EX and EE. Our findings show that AI-assisted applications for HRM enhance EX and, thus, EE. We also see increases in employee productivity and HR function's effectiveness. Implications for research and practice are also discussed.


| INTRODUCTION
With increased digitalization at the workplace, especially as we manage HR ecosystems, the use of artificial intelligence (AI) and other disruptive technology platforms for enhancing the employee experience (EX) of HR practices and employee engagement (EE) at work is gaining importance (Bersin et al., 2017;IBM & Globoforce, 2016). As workplace digitalization increases, employees interact with the various digitalized assets in an organization's ecosystem, yet, the ability of HRM to cope with the challenges of an AI-assisted HRM remains a concern Charlwood & Guenole, 2022;Swart et al., 2020). We know from prior research that introducing AI-assisted HRM applications in an organization's ecosystem creates anxieties among employees and can adversely (Presbitero & Teng-Calleja, 2022;Suseno et al., 2022) or positively affect employee and business outcomes . Nevertheless, little research examines the impact of AI adoption in HRM on EX and EE (Braganza et al., 2021). Despite its increase, there is a limited theoretical basis for understanding how AI-assisted HRM infrastructure fits into an organization's broader ecosystem and how firms design and implement an HR ecosystem and a configuration of digitalized AI applications to cater to the firm's digital, human and physical aspects of the work environment for improving EX and EE levels. Examining the adoption of AI-assisted HRM in a firm's ecosystem is also critical, especially as firms have faced EX and EE issues since the pandemic began (Hancock & Schaninger, 2020). Theoretically, this problem is further compounded as there is significant diversity in the conceptualizations of EX (Bersin et al., 2017;IBM & Globoforce, 2016;Malik, De Silva, et al., 2021;Maylett & Wride, 2017;Morgan, 2017;Plaskoff, 2017) and EE (Gruman & Saks, 2011;Kahn, 1990;Macey & Bunchapattanasakda, 2019) and the approaches employed (MacCormick et al., 2012;Mirvis, 2012). The above inconsistencies lead to problems of measurement and validity for advancing scholarly work. Though the theory of EE is well established, the term EX is often used interchangeably to misrepresent EE. Therefore, this article reviews the literature on EX, delineates the differences between the two concepts and suggests the relationship by offering an integrated definition of EX and its conceptualization. This is an important contribution to the literature on EX, especially in integrating AI-assisted HRM into an organization's ecosystem. It is critical to clarify how EX in the AI-assisted HRM applications in a firm's ecosystem critically influences EE (Shenoy & Uchil, 2018). Our review and empirical data confirm that EX is an antecedent of EE, thus allowing further research on the topic.
AI-assisted HRM applications can capture continuous and realtime employee data and perceptions from all aspects of an employee's physical, human, and digital work environment. This view is akin to the emerging literature on customer engagement platforms (Breidbach et al., 2014) and a recent conceptualization of customer experience (Becker & Jaakkola, 2020). Therefore, understanding how firms develop a configuration of AI-assisted HRM applications and platforms and fit it as part of their broader ecosystem is timely and needed for delivering the right set of EX and EE outcomes. Further, the strategic choices and knowledge regarding the configuration and quality of AI-assisted HRM applications designed and implemented are needed in the domain of HRM to achieve high levels of EX and EE. This approach highlights how HR leaders can exercise strategic choices to create a digitalized HR ecosystem for higher levels of EX and EE. Thus, based on the above, this research aims to establish how AI-assisted HRM fits into an organization's ecosystem? A related subaim is to investigate the impact of such AI-assisted HRM applications on EX and EE?
To summarize, this study offers the following distinctive contributions. First, it contributes by showing how AI-assisted HRM fits into the literature on HR ecosystems. Second, the impact of a firm's digitalized HR ecosystem comprising a configuration of AI-assisted HRM platforms on EE and EX (Gheidar & ShamiZanjani, 2020;Kim & Gatling, 2018;Shivakumar, 2020). Third, it clarifies the ongoing debates in the literature on EX and EE in using an AI-assisted technology-mediated social exchange. Finally, we develop an integrated theoretical framework for understanding how a configuration of AI-assisted HR applications fits into a firm's HR ecosystem and impacts EX and EE. We do this by employing the theoretical lenses of AI-mediated social exchange theory, EE, and engagement platforms (Blau, 1964;Breidbach et al., 2014;Kahn, 1990;Ma & Brown, 2020;Macey & Schneider, 2008).
Our choice of the above theoretical frameworks is relevant as, firstly, social exchange theory (Blau, 1964) is a foundational theory for EX and EE programs, as it is built on the foundations of trust and exchange of value between an employee and its workplace practices.
Employees experience stimuli from various sources and will thus react and respond to each stimulus based on their perceptions. Second, the AI-mediated social exchange theory (Ma & Brown, 2020) is also timely as employees interact and share knowledge through technologicallymediated AI applications that are part of a firm's digital HRM ecosystem. Third, the nature of the technologically mediated social exchange through these AI-based technology platforms invariably invokes a norm of reciprocity between the HR designers and employees (Ma & Brown, 2020), which affects the quality of EX. Similarly, the literature on EE (Bakker & Albrecht, 2018;Burnett & Lisk, 2019;Kahn, 1990;Smith, 2019) and the use of platforms that can enhance EX (Bersin et al., 2017;Burrell & Gherson, 2018;Shivakumar, 2020) is vital in understanding how and why firms exercise strategic choices in developing their digitalized HR ecosystem of AI-assisted HRM applications.
This article develops a theoretical framework and explains how firms can design and deliver excellent EX using various AI-assisted HR applications in their HR ecosystem by making certain strategic choices.
The rest of the article is organized as follows. First, a literature review on EX and EE, including social exchange theory and its new variant of AI-mediated social exchange, is offered (Blau, 1964;Ma & Brown, 2020). Second, literature on digitalized EE platforms and AI applications (Breidbach et al., 2014;Gheidar & ShamiZanjani, 2020;Kim & Gatling, 2018;Shivakumar, 2020) in studying the missing link -EXin influencing EE is presented. Third, the research methodology employed, analysis and findings follow. Finally, a discussion of the findings and a conclusion with implications for theory and practice.

| EX AND EE-A LITERATURE REVIEW
Extant literature suggests that a positive EX leads to higher levels of EE (Itam & Ghosh, 2020;Morgan, 2017;Plaskoff, 2017) and positive employee and business outcomes, such as work performance, discretionary effort and retention (IBM & Globoforce, 2016). However, a significant trend affecting work design and HR policies is the technologically savvy multi-generational workforce that demands a new way of thinking about their EX (Bersin et al., 2017;Burrell & Gherson, 2018;IBM & Globoforce, 2016;Shivakumar, 2020) and EE at the workplace (Kahn, 1990;Macey & Schneider, 2008). This approach departs from the current episodic and event-based system of EE (Budhwar & Bhatnagar, 2007) and exclusive talent management focusing on an exclusive rewards and benefits menu for key talent (Cappelli, 2008;Farndale et al., 2020;Scullion & Collings, 2011). Further, in large firms, new technologies can enable effective EE and management of diverse generational talents, using an AI-assisted HRM engagement platform for managing people in a firm's HR ecosystem (Malik, De Silva, et al., 2021). The review begins by explaining the differences between EE and EX and then presents a review, definition and reconceptualization of EX, positing it is a vital and missing link for high levels of EE.

| EXthe missing link for EE
Employee Engagement. In his seminal ethnographic work, Kahn (1990: p. 694), inspiring much of the scientific work on engagement, stated engagement as "harnessing of organization members" selves to their work roles; by which they employ and express themselves physically, emotionally, and cognitively during role'. The seminal work by Kahn (1990) conceptualizes EE as the psychological conditions (of meaningfulness, safety, and availability) of an individual's engagement or disengagement at work. Employees express themselves cognitively, emotionally and physically in their tasks when engaged. Therefore, this approach highlights the exercise of rational choice and individual agency on the part of employees (Cole et al., 2012). Others, such as Macey and Schneider (2008), used three different types of engagement-psychological state, behavioral engagement, and trait engagement as critical elements of EE, whereas, Schaufeli et al. (2002) defined engagement as "a positive, fulfilling, work-related state of mind that is characterized by vigour, dedication, and absorption" (p. 74). In practice, however, the dominant approach to EE is often measured as an episodic, one-off event or a series of time-bound activities, as evidenced through popular surveys like Gallup's Q12 survey (Gallup, 2021 We differentiate and build a deeper understanding of EX by drawing upon service management and customer experience literature. The literature review on customer experience notes three main approaches in explaining customer experience formation concerning accessing a service: individual, contextual or a mix of both (Becker & Jaakkola, 2020;Lipkin, 2016). We focus on the first approach-the individual-wherein individuals may be active or passive participants with some control over the experience formation process. In this stream of literature, most studies focus on the stimulus-organismresponse (SOR) model (Mehrabian & Russell, 1974) or a sensationperception framework (Fechner, 1860(Fechner, /1966, which has its theoretical roots in the environmental and behavioral psychology. The basic thrust of this approach is that an individual receives stimuli from the external environment, which affects an individual's (organism's) cognitive and affective state and leads to a behavioral response.
Similarly, the sensation-perception framework (Fechner, 1860(Fechner, /1966 argues that external environmental sensations trigger an individual's internal perception process, leading to a response. Both theoretical streams characterize perception as a process of interpreting responses through an individual's internal cognitive and affective state and attaching a meaning to the stimuli/sensations from the external environment (Goldstein & Cacciamani, 2021). Thus, perception is an intermediating mechanism between an individual's interactions and reactions to his/her external environment (Pareigis et al., 2012).
Building on the above concepts of customer experience management, Abhari et al. (2008) coined the term EX management, viewing employees as internal customers, and that firms must motivate their employees by delivering the right brand experiences and value-added offerings through five experiential needs of employees (cognitive, emotional, social, sensorial, and practical needs: p. 1). These experiential needs are well-documented in the literature on customer experience management (Becker & Jaakkola, 2020). Employees receive a vast range of stimuli from the external environment, and emotions are a generic term through which we express our experiences (Gordon, 1987in Harikkala-Laihinen, 2020. Emotional reactions are used to express our day-to-day experiences of the work environment.
So, as employees receive stimuli from the environment, after cognitive and affective processing of the stimuli, they express their experience via an intermediating process of perceptions and frame a response.
For example, an outcome of disengagement is triggered by poor experiences of sadness, despair or helplessness. The stimuli are received at multiple touchpoints in an employee's lifecycle, and these stimuli may not always be under complete managerial control; indeed, some stimuli may be coming from a dynamic interplay between work and nonwork domains. Depending on the quality of reactions and responses to these stimuli, this article argues that employee's reactions and responses to their physical, digital, and human aspects of work will affect their EX, which is why this article argues that EX is a critical antecedent (Plaskoff, 2017;Shenoy & Uchil, 2018) of EE (Kahn, 1990). Therefore, this article defines EX as continuous, non-deliberate, spontaneous, and real-time employee reactions and responses to diverse stimuli from the AI-assisted HRM applications in a firm's workplace ecosystem. We conceptualize EX based on three sub-dimensions: the stimuli' extent, nature, and perceived relevance. The extent of stimuli is the degree, strength, or frequency of stimuli employees perceive as available. The nature of stimuli is the diversity or range of stimuli and their perceived availability across multiple touchpoints. Finally, the perceived relevance of a stimuli captures, for example, employee motivation to respond to stimuli. See Figure 1 for a graphical model of EX. Several studies within the HRM literature have attempted to capture EX for specific areas of HR services, such as performance appraisal (Farndale & Kelliher, 2013) and flexible work arrangements (Chen & Fulmer, 2018). Often, the experience of such practices has been examined by characterizing employee perceptions. In this study, employees who receive stimuli from AI-assisted HRM applications and HR platforms would reasonably expect to associate it with a promise of increased social exchange with such AI-driven applications and an ongoing long-term exchange with the employee-organization relationship.
EX versus EE. Several differences exist between EX, a holistic, dynamic, longer-term, and employee-centric (Morgan, 2017;Plaskoff, 2017), and EE, an episodic, shorter-term, and organizationcentric approach. The literature on EE and EX suggests two distinct approaches-one where EE and EX are treated as unrelated concepts, and the second, where EE is viewed as a result of methodically planning EX. The former views EE as a set of actions and argues that EX affects EE. For example, if employees have a great experience at work because of the culture, technology, or physical spaces, they are much more likely to be engaged. Further, given the multitude of stimuli that employees receive from diverse sources at multiple touchpoints in their employment lifecycle, it makes it difficult for employees and their managers to process and analyze all their reactions, responses, and experiences due to issues, such as bounded rationality, memory recall or other organizational and individual constraints. EE is, therefore, a continuous and dynamic phenomenon and could change over time, based on EX-be it for a single transaction or several transactions. At a minimum, EX must be consistent and honor the EE's core tenet of a psychological contract as fulfilling a psychological contract antedates EE.

| Digitalized employee engagement platforms
Successful IT services firms have developed technology capabilities for leveraging business intelligence systems to respond to the diverse needs of internal and stakeholders by balancing multiple levels of strategic, horizontal, employee, and market fits (Breidbach et al., 2014;Boon et al., 2011;Malik et al., 2018). Moreover, with greater technological maturity and disruptions at work, the design of workplace and HRM practices are changing at both local and global levels (Lepak & Snell, 1999Morris et al., 2016), as evidenced by the debates on Industry 4.0 (Schwab, 2017) and increasing adoption of AI-enabled HR applications (Jaiswal et al., 2022;Vrontis et al., 2022). Consequently, digitalization and HR-focused AI applications at the workplace have freed employees' time to focus on non-routine and intrinsically rewarding tasks, thus increasing their overall EX of HRM practices (Malik et al., 2020).
However, some scholars have noted this as an attempt by HR to establish its legitimacy (Mahadevan & Schmitz, 2020). In addition, increased digitalization offers greater scope for integration and convergence of diverse stakeholder needs. Therefore, firms need to revisit their work and employment practice architecture to deliver sustained performance (Lepak & Snell, 1998;Snell & Morris, 2019;Swart & Kinnie, 2014) and EX through such platforms (IBM & Globoforce, 2016;Shivakumar, 2020).
With the increasing digitalization of EX touchpoints, EX of HRM practices in the physical, digital, and cultural domains through an AImediated social exchange is relatively unexplored (Malik et al., 2020;Malik, De Silva, et al., 2021;. EE and EX approach in a digitalized workplace are rapidly changing (Breidbach et al., 2014;Kim & Gatling, 2018). The literature on virtual and AI-based EE and EX platforms notes positive results from using such platforms. Accenture (2018), for example, found that companies with highly engaged workforces are 21% more profitable than those with low engagement.
This necessitates consideration of AI service quality and assimilation of the HR-assisted AI applications Prikshat et al., 2021) to effectively monitor, track, and analyze employee reactions and responses from diverse sets of static and interactional stimuli employees receive from their environment, including interactions with such AI-assisted HRM applications. In addition, well-designed AI applications can offer personalized and hyperpersonalized EX through AI-mediated social exchanges (Ma & Brown, 2020;Malik et al., 2020;Malik, De Silva, et al., 2021). This research suggests that the AI service quality and the perceived ease of using such AI-enabled applications positively impact EX and employees' usage intentions, thus, resulting in higher engagement levels. In addition, the use of AI-based applications for HRM practices has been linked to greater levels of personalization in learning and development (Whiteside, 2019), employee coaching (Barney, 2018), managing performance (Basu Mallick, 2019), administrative tasks (Haak, 2019) and other employee outcomes and HR effectiveness (Malik et al., 2020;Malik, De Silva, et al., 2021).

| Human and technologically-mediated social exchange theory
Building on the social exchange theory's key arguments (Blau, 1964), there are three fundamental aspects: relationship, reciprocity, and F I G U R E 1 Key dimensions of EX. *Stimuli are characterized by AI-assisted HRM Applications exchange. In a human-to-human social exchange, following a workplace relationship, the exchange between two parties typically relates to an economic or social value exchange between the parties involved (Blau, 1964). The norm of reciprocity underpins it. Further, with the increasing adoption of AI applications in digitalized workplaces, an AImediated social exchange typically occurs between employees or humans with other humans through machines (e.g., an AI application or a bot). The scholarship on the use of AI-based HRM applications and experience of AI-based HRM ecosystems is on the rise and highlights a positive impact on EX and HR outcomes (Makarius et al., 2020;Malik et al., 2020;Malik, De Silva, et al., 2021;Shivakumar, 2020). AI autonomous applications can act human-like and use their agency in a technologically mediated social exchange (Ma & Brown, 2020). At the centre of this technology-mediated social exchange is the value derived by the individual or the firm through such an exchange, such that a positive exchange, resulting in productivity gains or a positive EX can lead to a range of employee outcomes, such as increased intention to use AI applications, high levels of engagement and lower intentions to quit (Kim & Gatling, 2018;Malik et al., 2020;Malik, De Silva, et al., 2021). The employeeorganization relationship is an overarching term describing the relationship between the employee and the organization and includes micro-attachments such as the concepts of EE, psychological empowerment, and the psychological contract (Coyle-Shapiro & Shore, 2007). The exchange breaks down if there is a drop in such experience by either beneficiary involved. Thus, the AI-mediated social exchange is different from the traditional social exchange. See Table 1 below for differences between the traditional forms of social exchange and an AI-mediated social exchange. Unlike the physical forms of social exchange, an AI-mediated social exchange involves human interactions that are mediated through AI applications. The quality and extent of data democratization afforded by such AI-assisted applications increase the opportunities for employees' real-time, continuous, and hyper-personalized stimuli. This opens up opportunities for employees' rich and relevant reactions and responses, thus affecting their EX.

| METHODOLOGY
As this research explores a relatively under-researched topic of using AI-assisted HRM in a firm's ecosystem, capturing diverse sets of EX through EE platforms, our choice of an in-depth single case qualitative design of a large, unusually representative, and revelatory MNE is appropriate (Eisenhardt, 1989;Yin, 2003). Moreover, the choice of a single, in-depth case study design can illuminate a new phenomenon if researchers have access to a diverse set of informants and can collect rich data from multiple sources for triangulation and trustworthiness of the findings (Pratt, Kaplan, & Whittington, 2020;Pratt, Sonenshein, & Feldman, 2020).

| Indian research context
We chose India for its extreme ethnic, political, cultural, business, and spiritual diversity Mishra & Varma, 2019). It has a rich cultural heritage of ancient administrative guidelines of the Arthasharata dating back to 3000 BC, wherein the fundamental principles of holistic management practised in the past are still relevant to new knowledge-intensive firms in the Pharmaceutical, Ayurveda, healthcare, and IT industries (Malik et al., 2018;, including those at the bottom-of-the-pyramid (Basu et al., 2021). India is also home to the most highly sought-after technology talent globally (Pereira & Malik, 2015) and is indeed one of the most preferred geographical locations for MNEs seeking to establish their global innovation hubs in India Pereira & Malik, 2015). Even though there is an acknowledgment of India's deep pockets of technical and highly skilled innovative talent, there is still a dark side and cultural issues that need recognition and attention to give due credit to the innovativeness of Indian knowledge workers; else they end up being affected by neo-colonial power excesses (Malik, Mahadevan, et al., 2021). The differences in values between Generation Y expatriates and Generation X managers have been problematic for mGen Y employees (Pereira et al., 2017). The Indian IT industry has employed innovative HR practices to offer a unique expatriation experience to Gen Y expatriates to deal with such differences and manage and retain a vast population of millennials in the workforce (Pereira et al., 2017). Shenoy and Uchil (2018)

| About the case organization
To maintain the confidentiality and anonymity of the MNE's subsidiary operation in India, we use a pseudonym and refer to the MNE as BigTech Services. Pseudonyms are also used to describe BigTech's AI applications in its digitalized HR ecosystem. It employs over 100,000 highly skilled people serving more than 50 industries globally, with several ecosystem partners in digital, cloud, security, data and analytics, automation, AI, blockchain and industry X.0, and IoT technologies to offer immersive experiences and economic value and productive outcomes to its clients. Close to 90% of employees at BigTech Services are millennials and Gen Z and are a major driving force for Big-

| Data collection
The study's design involved questions to a vertical slice of the employees, managers and leaders (e.g., See interviewee details in  Feldman, 2020;Yin, 2003). The nature of the questions followed a semi-structured approach and included, for example, the following.
First, the questions focused on the nature and extent of technological innovation designed and implemented by BigTech in AI technologies in various functional domains. Next, this was followed by focusing on how these applications were designed and implemented for their HRM function. The next set of questions examined how it affected the business and end-users (e.g., managers and employees) experience of these AIassisted HRM applications. As part of this, the HR managers, designers of the AI-assisted solutions, and employees were asked further questions about their experiences of the HRM practices through these applications and with the broader AI-assisted workplace ecosystem comprising physical and digital artifacts. Some illustrative questions asked were: How clear are you about your goals and objectives? When was the last time you got recognized for your contributions? What promotes (and prevents) collaboration for you at the workplace? Are you able to speak without any fear of retribution? -these questions were aimed to elicit EX across various HRM processes. Finally, a deeper probe of the employees and managers focused on questions related to the end-users experience at an employee outcome level (e.g., satisfaction, loyalty, and commitment). These questions also examined the efficiency and effectiveness at a business and strategy level.
In addition to the abovementioned interviews, the nature of documents and other sources of data analyzed include: non-participant observations, interactive communications with the AI applications, organizational records, client case studies and testimonials, HR policies, user satisfaction data, leadership, and strategic values framework, post-hoc Q12 EE reports, attrition reports, performance management analytics, code of business ethics surveys, staffing reports (which helped us understand the pattern of skill adoption and skill inventory), culture surveys, candidate experience analytics, and employees' exit interviews. Additional secondary data were accessed through its website. The use of additional data sources allowed for data triangulation and supplemented the themes analyzed. For example, end-user surveys of employees and frontline managers on using these applications and reports focusing on the PeopleXp framework were highly beneficial in corroborating evidence from more than one data source. The number of interviews (23, in this case) is vital for data saturation (Saunders & Townsend, 2016), which occurred in the first 14-15 inter-

| Data analysis
Interview data, organizational documents, white papers, HR policy framework, clients' case studies, and publically available information from the MNE's website were analyzed. More than 125,000 words, non-participant observation, and informal communications with AI applications formed the total pool of data analyzed. Access to these rich sources of case study data facilitated a deeper understanding of EX's contextual influences and employees' perceptions and interactions with an AI-mediated social exchange.
Following transcription of interview data and collation of additional documents, concepts, and themes were developed using traditional manual coding approaches. The manual coding generated the following first-order codes: people, teams, rewards, projects, work, diversity, needs, knowledge, innovation, technology, ideas, experience, voice, solutions, and problem-solving. This was followed by coding a targeted theoretical second-order coding of theme, such as values, HRM practices, physical, digital, and values ecosystem, and EX and EE, using a theoretically informed abductive logic approach (Dubois & Gadde, 2002;Van Maanen et al., 2007) for our analysis. A theoretical coding process followed the preliminary coding of the first-order codes from the raw textual data. Abductive theorizing requires going back and forth between the data, relevant theories, and the case phenomenon. Where it was necessary to resolve an emerging puzzle from the phenomenon regarding the relationships or new concepts, a constant interplay between data, theory (Blau, 1964;Breidbach et al., 2014;Kahn, 1990;Macey & Schneider, 2008;Ma & Brown, 2020), and phenomenon was employed. Theoretical coding was undertaken using theories such as HRM, AI adoption, EX, and engagement. Next, this theoretical coding led to developing our theoretical framework and mapping the nature and extent of EX and its diversity ( Figure 2).

| ANALYSIS AND FINDINGS
This section begins with how BigTech designed AI-assisted HRM to improve EX and EE outcomes in its ecosystem. Next, an analysis of four meta themes is presented, and where applicable, integration of the findings with our theoretical model.

| INTEGRATING AN HRM PRACTICE ECOSYSTEM
As part of the human element of the PeopleXp framework (see  Technology Leader] for the cause. This generated an awareness of the drivers of employee health and wellbeing.

myRewards and recognition. Appropriate recognition and
rewards at the workplace can help extract an employee's discretionary efforts. However, for any reward or recognition initiative to be effective, it needs to be timely and visibly impact the employee's social currency. In other words, the recognition needs to be real-time, and

form. In addition, rewarding and recognizing talent occurs through its
Talent bot and is linked to a series of reward programs. myTalent and passion. Employees bring more than their professional competence to work. This is where managing work and nonwork stimuli for excellent EX comes into play. BigTech recognized the need to allow employees a social vent and a physical platform to express their other non-work talents, which greatly mattered to employees, thus increasing their EXs. Organizations need to identify, appreciate, and, more importantly, create a platform to recognize this "personal" side. The myTalent and passion theme and HR vertical helped employees bring their side to life through sports, culture, and performing arts initiatives. When employees discover that the person next to them is a state-level badminton player, has sung a popular song in a full-length commercial motion picture, or is an Olympics medal winner, the level of admiration goes up several notches. Thus, acknowledging and addressing non-work-related issues and providing support helps increase employee motivation and commitment (Mohamed et al., 2006

| LEVERAGING VALUES, DIGITAL, AND PHYSICAL ENABLERS
The PeopleXp framework is meant to engage with employees as it is designed collaboratively by business and HR teams with sponsorship from senior-most business and HR leaders at BigTech's Technology and business division. Each framework element is co-owned by a business lead and an HR professional to deliver its intended agenda.
Extensive intra-and inter-team collaboration focus on areas to synergize efforts. The business teams find immense value in work delivered through the three horizontal elements of the framework (myvalues and culture, mydigital and myphysical) using AI-based applications in the HR ecosystem. In addition, they provide a real-time focus through the AI applications in typical customer engagement and experience domains.  Table 3. Further analysis of this theme led to the development of our theoretical model (see Figure 2). Below are several AI-assisted applications that BigTech's digitalized HR ecosystem employed across the MNE's geographies, offering diverse EXs, and varied levels of AI-mediated social exchanges.

| THREE HORIZONTALS: MYVALUES AND CULTURE, MYPHYSICAL AND MYDIGITAL
Phygital. The Phygital platform was used to digitalise employees' physical ecosystem is a critical part of the overall HR ecosystem. Also, as part of the Phygital platform, having user-friendly cashless digital applications to order food at employees' convenience has been implemented across the Indian subsidiary. As a result, 100% of employees have downloaded this app. An extension of this was the digitally enabled hot beverage vending machines offering freshly brewed coffee and natural beverages free from preservatives and sugar (Malik, A., Personal Observation, 2019). Another major part of the Phygital platform was developing online taxi bookings, food ordering, and securing entertainment events. This was particularly relevant for showcasing events to employees, who exhibit a vital social need for connection and sharing their social skills and expertise with other colleagues within the social relations at work. This engagement is of high cultural value in an Indian setting. Thus, this app focused on creating workplace pilot designs that support flexible and fluid working methods while addressing the human need for social interaction.
Although these are just a few benefits, employees can experience an engaging workplace, conveying a sense of wellbeing and care.
Team Insights. A team of three functional experts, 10 leads from the field HR team, and two from the Analytics Centre of Excellence collaborated with 10 technical experts to design and deliver 60+ reports through their technology platform. This team created a data visualization tool called Team Insights, which integrates all HR analytics models in disparate HR sub-systems. It is built on the QlikView technology platform. With an exponential increase in employee headcount, the human resource function and its sub-functional domains, such as recruitment, workforce planning, rewards, training, engagement and retention, were integrated with this interface to deliver focus at scale. As the fulcrum of any business transformation lies in talent transformation, the business leaders must effectively plan their workforce. This dashboard provides insights into a report and correlates it with additional (and supporting) data to support managerial decision-making.
The  tools. IiPA has now answered over 1 million queries with >80% accuracy (Internal Document, 2019). Of 60% of users, more than 50% were repeat users. In the fully mature "Coach phase" of its evolution, IiPA will serve as a solution provider, making proactive recommendations on a wide range of employee-related matters-from compensation break-up and vacation planning to career guidance and wellness.
myCompetency and Talent Bot. BigTech Services defines competencies as the combination of knowledge, skill, and process abilities that an individual requires to perform a job. According to an internal survey conducted by BigTech Services with more than 10,000 employees worldwide, 64% of respondents agreed that the pace of change in their job is speeding up due to technological change (BigTech Survey, 2019). Thus, finding the right employee at the right time with the required skillset from a large workforce base is akin to finding a needle in a haystack. With over 100,000 employees and considerable skills to map, the problem becomes complex. Running with an anachronistic annual performance management cycle can lead to only episodic training/skill need identification once a year. Performance and skill/competency development in the current digitalized disruption requires continuous review. This perhaps is the only way to match the external demand and internal supply source. If the demand is elastic, it could be fatal to have plasticity in internal processes. It would also be imprudent to forever recruit staff through external hiring due to cost and productivity challenges-increasing the hiring cost, uncertainty in staffing a project, and high assimilation and lead-up times for recruits.  This platform's active use has led to more than 1600 unique users (not to count the repeat users) and greater than 90% effectiveness in query resolution (Internal Dashboard, 2019

| DISCUSSION
The findings point to high EX levels (Plaskoff, 2017;Malik et al., 2020;Malik, De Silva, et al., 2021;), which positively impact EE (Kahn, 1990;Shenoy & Uchil, 2018) through the use of AI-assisted applications. As part of an engagement platform, these applications offer ongoing personalization and hyper-personalization of EX of HRM practices at different touchpoints in the employment lifecycle. In addressing the study's overarching aim, the role of AI-assisted HRM in a firm's ecosystem reflects the strategic choices leaders exercised to retain differentiated talent groups and ensure high levels of the psychological contract between the organization and these groups.  (Malik, De Silva, et al., 2021), thus, EE, leading to improved employee commitment, retention, and business level productivity outcomes (Malik et al., 2020). BigTech's HR managers' and business leaders' choice of investing in digitalized HR platforms created a positive EX for all three underpinning horizontals of human, digital, and physical domains (Bersin et al., 2017;IBM & Globoforce, 2016;Morgan, 2017;Plaskoff, 2017). The high quality of EX and personalized care through these AI applications triggered a norm of reciprocity and trust (Malik et al., 2020;Malik, De Silva, et al., 2021), resulting in better engagement, loyalty, and commitment levels. This relationship was stronger as most of the employees at BigTech were co-creators, designers, and consumers of these AI applications. For example, the Phygital platform was entirely driven by employees' emotional and physical needs; consequently, the buy-in and satisfaction with its usage were very high.
Employees reciprocated through higher satisfaction and commitment levels as they adopted these workplace changes without resistance (Blau, 1964). They felt a sense of pride working at BigTech, and an innovative MNE brand's AI-mediated exchange (Ma & Brown, 2020) provided strong reciprocity (Becker & Jaakkola, 2020).
Our theoretical framework clarifies the purpose and nature of the AI-assisted HR ecosystem of EX and EE platforms (see Figure 2).

| IMPLICATIONS FOR HR AND BUSINESS TEAMS
We identify several implications of our theoretical framework for theory and practice. First, we note an expanded role of HR leaders and managers as co-designers of an AI-assisted digitalized HR ecosystem of multiple HRM, EX, and engagement platforms (Breidbach et al., 2014). For HR managers to implement a new AI-assisted HR ecosystem and platforms, they must have developed digital and data science skills (Malik et al., 2020). In addition, they should show a change in their mindset and reimagine work and how to manage and work with people as they offer their core services. Second, HR leaders and managers must identify and address employees' latent needs in From an HR ecosystem theorization perspective, a critical point is the purpose and nature of platforms in a firm's digitalized HR ecosystem. BigTech's HR ecosystem positioned itself well to deliver high EX levels (Kim & Gatling, 2018;Malik et al., 2020;Malik, De Silva, et al., 2021). In developing such an ecosystem, HR managers must cocreate applications by working collaboratively and with the direct involvement of a diverse group of stakeholders, including employees-to realize high levels of EX. Managers have an active communication role in ensuring employees have access to self-select menus with adequate flexibility in choosing a range of personalized and hyper-personalized HR solutions (Malik, De Silva, et al., 2021).
Managers can leverage employees' skills by encouraging collaboration and knowledge sharing through traditional and AI-mediated social exchanges (Malik, Nguyen, & Budhwar, 2022;. Evaluating reciprocity norms, employees assess their benefits from exceptional EX platform(s), and many work for an organizational brand that offers innovative technology-mediated exchanges (Ma & Brown, 2020). Finally, the new skills and HR competencies needed by employees and HR practitioners to continue to deliver value in digitalized HR ecosystems include digital savviness, data fluency, and technology coaching. In addition, working across disciplines and functional areas to collaborate and co-create and design AI applications requires new skills in working with cross-functional teams to design and deliver HR programs.
For organizations to succeed, their internal customers (employees) must have positive EX to reciprocate it by engaging and committing to their workplace for extended periods. Benefits would accrue to organizations as their employees are more satisfied and engaged with their workplace and have a lower intention to quit. The parameters to track would be attrition, revenue, profitability, and an employer of choice brand and employee and customer satisfaction at any predetermined/Adhoc point in time. The HR professional will need to make a gradual shift from having a process-specific (generalization) expertise through deep expertise (specialization) to personal mastery (personalization) and developing sound expertise in coaching (individualization). In addition to the skills noted below, these will be part of HR leaders' and managers' new and expanded role in supporting such an ecosystem.
Other competencies that HR practitioners will demand are digital savviness-a high technology quotient, data fluency from actionable insights, using advanced decision support systems, and undertaking an expanded coaching role. In addition, HR leaders and managers will require high emotional intelligence and creativity levels, a characteristic that is not the core of automation and digitalization yet. To align HR service delivery to the realities of the changing landscape, several new possibilities of automation for HR need exploring. The efficiencies that will be realized will reduce manual efforts and allow the HR team to focus on human skills-such as relationship building, critical problem solving, strategic thinking, and innovation.
EX, efficiency, and effectiveness are the driving forces underpinning a digitalized HR ecosystem. This will lead to leveraging automation efficiencies and evaluating effectiveness to provide individualized EX, an antecedent to EE. We also need a supportive company culture comprising collaborations, transparency, psychological safety, alignment, and sharing feedback. These elements collectively raise exit barriers and help create an ecosystem for employees to thrive and innovate. Further, looking at integrated facets rather than disparate focus areas can improve EX, characterized by personalization and individualization. Although personalization addresses the needs of personae, individualization makes it unique to an individual's needs.

| IMPLICATIONS FOR THEORY
The knowledge-sharing intentions and motivations when using an AImediated social exchange. This is critical as some actors in the ecosystem may be more skeptical of the advantages of an algorithmic system of AI-enabled HRM EX and EE platforms. A related question is whether such systems can create value for all actors in the ecosystem, or will only a few stand to benefit? As the boundaries between work and non-work domains become blurred, further research on managing non-work-related issues and talent on a range of employee and business outcomes is warranted (Mohamed et al., 2006

| CONCLUSION AND LIMITATIONS
In a supply-side labor market, employees are lured away by better wages, benefits, and cultural factors that encourage employee autonomy, well-being, and work-life balance. If the organizations can address these factors through the values, human, physical, and digital continuum, it augurs well for employees who see a lasting value in their association with the firm. Despite our novel findings, our study is not devoid of limitations. First, an in-depth single case study design from an emerging market context warrants further validation in other contextual settings. However, the generalizability of our study's findings is not to the wider population but the relevant theories employed. As our data were from a large technology MNE, our findings may be relevant to small IT service providers or larger domestic firms embarking on a digitalization journey of their HR ecosystem. Also, comparative case studies of advanced adopters of AI-enabled HR applications versus producers and consumers of AI firms would yield meaningful differences and insights into AIenabled HRM practices. Second, we could not factor in macroenvironmental factors such as the impact of major crises such as the pandemic or other significant events and how that might shape the digitalized HR ecosystems, a topic of further inquiry. Finally, our conceptualization does not investigate ethical challenges and the dark side of AI-enabled HR applications as part of a broader digitalized HR ecosystem. Future research can explore the impact of poor EX on adverse employee outcomes and ethical issues (Ma, 2019).
Finally, our findings are more applicable to firms co-creating such AI-based applications with their employees, and such an experience may not eventuate in user firms who are buyers of such applications from a disparate vendor base.

ACKNOWLEDGMENT
We thank and acknowledge three anonymous reviewers, special issue Librarians.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are not available due to the University's ethical requirements and agreements with the participating case organization and interviewees.