Preparing the Workforce for AI Adoption: How Mohammed Shahin Sees Reskilling and Responsible HR Strategy Supporting Digital-Economy Competitiveness

Artificial intelligence is changing the way organisations manage information, assess performance and organise work. Yet introducing an intelligent system is not the same as preparing a workforce to use it. Employees need opportunities to learn; managers need sound judgement about what technology can and cannot do; and HR departments must ensure that efficiency does not come at the expense of fairness. The organisations most capable of adapting to AI may be those that treat digital transformation as a workforce-management challenge rather than simply a software purchase.

For Mohammed Shahin, a senior human resources and administration professional with more than two decades of experience in Bangladesh, that distinction is especially relevant. His career combines long-standing experience in employee relations, performance management and workplace administration with postgraduate human resource training and a Master of Science in Information Technology. Together, these qualifications offer a useful lens for considering how organisations can modernise their operations while retaining the human processes on which effective workplaces depend.

Two decades at the intersection of people and operations

Shahin worked at Stylecraft Limited in Gazipur from June 2001 until December 2021, progressing to Senior Deputy Manager – HR & Administration. His responsibilities and professional scope included employee relations, grievance handling, disciplinary procedures, departmental enquiry, performance management, HR documentation and administrative coordination. Such work involves more than enforcing policies. It requires consistent procedures, communication across departments and an understanding of how management decisions affect employees.

These capabilities become particularly significant when employers introduce AI-assisted recruitment, workforce analytics or automated scheduling. A system may process information quickly, but employees still need to understand why a decision was made and whom to approach if it appears inaccurate. Existing grievance procedures, supervisory responsibilities and documentation practices should not disappear when a decision includes an algorithmic recommendation. They may need to become more explicit.

Reskilling must begin with the work itself

AI training is often reduced to a one-off workshop. A more effective approach begins by identifying the actual tasks likely to change. Administrative staff may need to learn how to verify machine-generated summaries; HR personnel may need to assess automated screening outputs; supervisors may need to recognise when a recommendation should be challenged. This task-based view separates broad AI awareness from the deeper training required for specific roles.

Employers can then map existing capabilities against new requirements, provide opportunities to practise in realistic situations and evaluate whether training improves work quality. Employees who already possess valuable institutional knowledge should not be overlooked simply because they are unfamiliar with new tools. Reskilling can build on that experience, combining domain knowledge with digital literacy rather than assuming technology makes prior expertise obsolete.

Employee trust is part of implementation

Technology adoption can generate uncertainty about job security, surveillance and performance evaluation. That uncertainty may intensify when organisations introduce systems without explaining their intended uses or limitations. HR teams can help by establishing clear communications, accessible channels for questions and procedures for escalating concerns. Employee involvement during pilot programmes can reveal practical problems that technology teams might otherwise miss.

Shahin completed specialist training in disciplinary action, grievance handling and departmental enquiry through the Bangladesh Employers’ Federation in 2017. His professional development also includes training in Bangladesh labour law and industrial relations, and in human resource management and social compliance. These credentials are relevant to the governance side of AI adoption: organisations need to consider not only whether a tool functions, but whether its use is compatible with workplace obligations, transparent processes and respectful treatment.

Responsible AI requires human accountability

AI systems can support HR activities, but automated outputs are not automatically objective or complete. Data may be outdated, unrepresentative or collected for a different purpose. A screening model can overlook relevant experience; an employee-performance dashboard can miss context; a scheduling tool can optimise coverage without adequately considering worker needs. Responsible implementation therefore calls for documented purposes, proportionate data use, regular review and meaningful human oversight for consequential decisions.

This does not mean avoiding automation. It means distinguishing low-risk administrative assistance from decisions that may materially affect hiring, promotion, discipline or employment conditions. Organisations should clarify who is accountable for reviewing results, how errors are corrected and how employees can seek explanations. These safeguards are essential to maintaining confidence in both management and the technology it adopts.

A career shaped by continuing education

Shahin’s professional development reflects the increasing overlap between management and technology. He earned a CEMBA from Bangladesh Open University in 2013, a Postgraduate Diploma in Human Resource Management from the Bangladesh Institute of Management in 2018 and an MSc in Information Technology from St. Francis College in New York in 2024. He has also completed training in performance management, SME management and office management. This progression brings operational experience together with formal study across business, HR and IT.

His training in performance management is particularly pertinent. AI-enabled measurement can produce more indicators, but a larger volume of metrics does not necessarily improve judgement. Effective evaluation requires relevant standards, opportunities for feedback and an understanding of circumstances that raw figures cannot fully represent. Human resource strategy should connect new analytical capabilities to employee development, not allow measurement to become an end in itself.

From AI ambition to practical workforce strategy

A realistic organisational roadmap can start with a review of existing workflows and a skills-gap assessment. Employers can then select limited use cases, establish data and oversight safeguards, and provide role-specific training. During pilot implementation, they should gather feedback from employees and managers, track errors as well as efficiency gains and revise procedures before scaling. HR, IT, operations and legal or compliance functions all have roles to play.

The objective is not simply to make employees faster. It is to develop an adaptable workforce able to use technology appropriately, recognise its weaknesses and maintain service quality. For businesses competing in increasingly digital markets, these capabilities can support continuity, productivity and organisational resilience. They also reduce the risk of investing in systems that staff cannot confidently or responsibly use.

Shahin’s professional profile illustrates why workforce transformation benefits from both technological awareness and established HR expertise. His career does not document a specific AI deployment or a published study of AI reskilling, and the arguments here should be understood as an analysis informed by his documented background rather than a report of his original research findings. Nevertheless, his combination of long-term people-management experience and information technology education speaks directly to a central challenge of the digital economy: organisations will only realise the value of AI when they prepare, involve and protect the people expected to work with it.