AI-Powered Onboarding

Interview with Mirko Puliafito

Does AI replace part of HR management or does it integrate and enhance it (or both)? What approach would you suggest for assessing the contribution of AI technologies to HR roles and processes?

AI is not replacing HR management, but rather transforming and enhancing it by automating repetitive tasks and leaving MORE room for the HUMAN and STRATEGIC ASPECTS of the function.

We can think of HR processes as a continuum, from operational activities (CV screening, onboarding, administrative management) to more strategic ones (talent development, engagement, culture).

AI can intervene at both levels, but with different effects:

IT AUTOMATES low-value tasks such as screening candidates, managing documents or answering internal FAQs.
IT SUPPORTS and ENHANCES strategic activities by providing data-driven insights, predictive analytics and tools to personalise employee support.

So, AI is an ally, not a replacement. It can help make HR work more effective and focused, but it cannot (and should not) eliminate the role of empathy, negotiation, and human understanding that define the function.

How did you get involved in AI for HR processes?

I got involved in AI for HR processes because of a specific need: to make onboarding and training paths more effective and personalised, overcoming the rigidity of traditional systems. I have done this in business and, in the past, in technology by introducing AI into the Docebo LMS platform.

During my journey, I realised how difficult it is to ensure a truly tailored onboarding experience for each individual. HR processes often rely on standardised rules, checklists and one-size-fits-all documents, leaving little room for dynamic and adaptive interaction.

AI has allowed me to transform this approach: from a one-way model to a system that LISTENS, LEARNS AND RESPONDS TO THE NEEDS OF EACH PERSON IN REAL TIME. Thanks to data analysis and advanced interaction tools, we can now create more INTELLIGENT, EDUCATIVE and ENGAGING onboarding paths, where technology serves people and not the other way around.

At Digitiamo, we believe that AI should not replace the human touch, but enhance its impact. We work to build solutions that help people grow, integrate and feel part of an organisation from day one.

In your vision, what elements should be considered when talking about onboarding?

Onboarding encompasses several elements that HR today addresses in varying degrees: hard and soft skills, alignment with expectations, training on internal company procedures, training on company specifics, workplace safety training, individual assessments, goal setting, and possibly ongoing monitoring of the person’s performance in the role.

In summary, the key elements are ASSESSMENT, TRAINING and ADAPTATION to the EXPECTATIONS of the ROLE to be filled and to the CULTURE of the company.

Onboarding therefore addresses a need for LEARNING and a need I call POST-ASSESSMENT.

What are the main challenges of onboarding NEW RESOURCES compared to an internal career improvement process?

For internal transfers, onboarding can focus on training for the new role, whereas for external hires, there is greater UNCERTAINTY about the person’s skills and adaptability, so – in general – a more thorough assessment and onboarding plan are needed.

In its traditional sense, the concept of onboarding refers to the moment when a new person joins the company. How do you define it?

In our vision, onboarding is not just the first step of a new employee in the company, it is a necessary process in different scenarios and when a new hire or career advancement occurs, it ensures that the person becomes active and autonomous in covering a specific role.

Onboarding is therefore a CONTINUOUS PROCESS OF INTEGRATION, LEARNING, AND ADAPTATION.

Traditionally, it is seen as a set of activities to introduce a new resource to the company’s culture, procedures, and colleagues. But in my view, it is a dynamic ecosystem that doesn’t end in the first few days or weeks but EXTENDS OVER TIME TO ENSURE THAT THE PERSON IS TRULY AUTONOMOUS AND PERFORMING in their role.

Onboarding is more than just a welcome; it is an ACCELERATOR OF EFFECTIVENESS.

In the modern context, we can distinguish three levels of onboarding:

Functional Onboarding – Provides the tools, information, and training needed to get started.
Cultural Onboarding – Promotes alignment with the company’s values, vision, and mission.
Relational Onboarding – Builds connections with the team and the company network to accelerate integration.

At Digitiamo, we see onboarding as an ADAPTIVE AND PERSONALIZED PROCESS, that addresses different scenarios and where AI plays a crucial role in ensuring that each person receives the most appropriate information and support tailored to their profile and needs.

The ultimate goal is not just to bring someone into the company but to enable them to perform at their best in the shortest possible time and with the best possible experience.

L’onboarding è più di un semplice benvenuto, è un acceleratore di efficacia.

Nel contesto moderno, possiamo distinguere tre livelli di onboarding:

Onboarding funzionale – Fornisce gli strumenti, le informazioni e la formazione necessarie per iniziare a lavorare
Onboarding culturale – Favorisce l’allineamento con i valori, la visione e la mission aziendale
Onboarding relazionale – Costruisce connessioni con il team e il network aziendale per accelerare l’integrazione

In Digitiamo, vediamo l’onboarding come un processo adattivo e personalizzato, che riguarda diversi scenari e in cui l’IA gioca un ruolo fondamentale nel garantire che ogni persona riceva le informazioni e il supporto più adatti al proprio profilo e alle proprie esigenze.

L’obiettivo finale è quindi non solo far entrare qualcuno in azienda, ma metterlo nelle condizioni di dare il meglio di sé, nel minor tempo possibile e con la migliore esperienza possibile.

Do you see this as a complex process that can be effectively enhanced by AI?

Absolutely, yes.

It is not a static and one-size-fits-all process, but needs to be STRUCTURED, FLEXIBLE AND PERSONALIZED, based on the role, previous experience and goals of the individual.

STRUCTURED because it involves several dimensions: knowledge of the company, technical training, cultural integration, development of soft skills, and building internal relationships. Effective onboarding doesn’t just provide information, it ensures that THE PERSON FEELS LIKE AN ACTIVE PART of the company.

PERSONALIZED because everyone has a different background, experience, and learning style. A junior profile needs different support than a senior, just as a technical profile has different needs than a commercial one.

How does AI improve this?

AI allows to SCALE PERSONALIZATION and EFFECTIVENESS of onboarding in several ways:

CUSTOMIZED PATHS: It analyses skills and needs to offer targeted training content

FAST ACCESS TO INFORMATION: Chatbots and digital twins answer questions in real-time, eliminating the need for lengthy document searches

MONITORING ADAPTATION: It analyses the individual’s engagement and progress, highlighting potential difficulties before they become problems.

CONTINUOUS SUPPORT: Instead of a “time-limited” onboarding (e.g., the first two weeks), it can ensure intelligent mentoring during the first few months of activity.

You mentioned not assessment but POST-ASSESSMENT. Why?

I talk about post-assessment because after selecting a person, since there is no such thing like the perfect resource, there will definitely be a positive background that the person brings and for which they were selected, but there will probably also be some GAPS TO BE FILLED.

This is always true, but especially in the case of a REPLACEMENT, because there will be expectations regarding the new resource, both in terms of the business context and the specific role they will be taking on.

I would add that – jus as importantly – AI can have an impact on the ENGAGEMENT of the new resource, precisely because it pays attention that is truly tailored to the individual.

How could AI make a difference, in this case?

AI allows us to improve what we already do in response to these needs, primarily because it can provide a much more accurate ASSESSMENT of the person in terms of skills, but also of values, personality traits, and interests, compared to what was done during the selection phase.

You mentioned that when examining the onboarding process, it’s important to consider different scenarios when looking at the onboarding process. Which ones?

We think it’s useful to distinguish three specific cases of onboarding, as each has different dynamics and challenges and requires different approaches, processes, and tools. Specifically:

New person | New role (with which the company has no previous experience)
Internal person | New role (career improvement)
New person | Existing role in the company (replacement)

AI can contribute in a targeted way to make onboarding more effective in each of these cases.

Let’s analyse them one by one:

New person | New role

How to implement effective onboarding for a role where the company has no experience?

If the role is completely new (for example, the creation of a marketing service that was previously outsourced), there will be greater RELIANCE on the NEW PERSON and on what was learned about them during the selection phase.

Even the creation of the new role will be based on what the new person has done in the past.

In this case, the need for a post-selection assessment seems less urgent and less controllable: since the role is new, there is no internal reference to challenge, verify or double-check; and even if a post-selection assessment were carried out, the awareness of what the person lacks would still be relatively low.

Let me give you an example “from my own experience”: I decided to introduce a new sales manager role and chose someone based on a network of references, the people who spoke to me about them and the head-hunter who assisted us. But in reality, the person lacked alignment with the company values, as well as lacking relationship skills, deep-dive skills and a number of other things that we only realised after subsequent assessments.

It was the first role of its kind in the company, we had no experience and, looking back, I probably should have done a lot of things differently, identified the gaps much earlier, either during the selection phase or during onboarding.

Also, you can’t expect to teach a new person a role that you don’t know yourself: there is an UNKNOWN, you don’t know exactly what the person will do in that role, so you don’t know how to integrate their skills properly.

So, unless you use external consultants – or perhaps the same head-hunter who helped you recruit – your only option is to wait for the results.

In addition, if you can’t rely on the resource directly because they are highly skilled or because they need to deliver results in their role immediately, it may also be appropriate to have an external coach to help and support the person alongside the company’s onboarding process.

What else can AI do to create an effective onboarding for a new resource in a role where the company has no experience?

In this case, AI is incredibly effective in improving onboarding and can make a real difference.

Because while the company internally may not know exactly what is right or wrong for a new role, AI acts as a third eye that – instead – has the experience of all those who have already held the same type of role and have also built learning paths accordingly.

This is where AI helps us: while HR tends to focus on the people in the organisation, AI fortunately has the EXPERIENCE OF OTHERS that it inherits and makes available.

When our assessment skills seem inadequate, you can rely on a tool that is open and experienced.

Internal person | New role

What are the benefits and challenges of CAREER IMPROVEMENT?

We can think of CAREER IMPROVEMENT as ‘disguised’ onboarding: the person hasn’t joined the company as a new employee, they’ve just changed roles. If someone is promoted internally, there will most likely be a MIND THE GAP path. However, there will obviously be no training on the company itself.

There needs to be a learning path that starts with an assessment of the person’s skills against those required by the new role. So, it is a UPSKILLING ORIENTED TOWARDS CAREER IMPROVEMENT.

For example, we developed someone from a Marketing Specialist to a BDR (Business Development Representative), so that they were not just posting on social media, but also doing outreach. We created a training path to evolve their profile from A to B, providing them with all the tools for “how to make cold calls, where to find the lists, how to approach, how to create an agent script, how to manage dialogues with others.” These are all typical onboarding upskilling activities.

By the way: career improvement does not always involve upskilling. For example, in the case of an internal replacement of an existing role: if the CMO leaves and I replace him with someone from his team who is more junior but already has certain skills, it doesn’t necessarily require a lot of upskilling.

There is also completely free upskilling: when I used to work with LXP (Learning Experience Platform) systems, the resource was free to train on new technologies, for example.

How can AI improve upskilling?

AI can improve upskilling by personalising learning paths based on a person’s current skills and goals.

By analysing performance data and learning preferences, it can suggest targeted content, adjust the difficulty of exercises and provide real-time feedback.

In addition, AI-based tools can act as virtual coaches, providing hands-on simulations, role-plays and contextual advice.

This allows for more seamless professional development that meets both individual and organisational needs.

NEW PERSON | EXISTING COMPANY ROLE

What are the challenges of REPLACING an existing company role with a new person?

When I need to replace an existing role – because the person has left and I haven’t found an internal replacement – with a new external resource (or even when I add an external resource to an existing team), we have the most complex case, which combines the two previous ones.

Once you’ve selected a new resource, you expect the new person to bring the same type of contribution as the person he or she is replacing. This person has – positively or negatively – consolidated and accustomed the company to a certain type of contribution, work and relationship.

In the case of a replacement, therefore, there is always an EXPECTATION OF THE ROLE that must be taken into account.

Moreover, beyond the expectations related to a specific role, it is certain that a new person in the company cannot be expected to contribute in the same way as the person who was already operating in that context.

Therefore, from an onboarding point of view, it’s important to make an ASSESSMENT (which should be done during the same onboarding phase or earlier if a screening has been carried out during selection and we know what the person lacks) in order to build a LEARNING PATH to fill the GAPS that the new person will certainly have compared to the previous employee.

This will allow the person to be equipped with the necessary tools to become available, active and autonomous as soon as possible, without the need for on-the-job training.

It would not be possible to plan on-the-job training in advance, as the missing skills would only become apparent after the assessment.

In addition, coaching intervention may also be necessary, especially in the case of empowering the internal team.

How can AI improve the replacement process?

AI helps to manage the replacement process, both during the selection phase and during the integration of the new resource.

It can identify gaps compared to the previous person and suggest personalised onboarding paths to reduce adjustment time.

In addition, by continuously monitoring performance and analysing team interactions, AI can highlight potential issues in the transition process and facilitate more effective alignment between the new resource and the corporate culture.

How can AI help companies optimise the onboarding process?

AI enables the creation of training content and provides only the necessary information and training for a specific role.

Let’s say that every onboarding requires the company to have

. policies that describe its life and rules

. an orientation path for each specific profile (compared to what has been published in job advertisements or, in any case, according to the needs for which a resource has been added).

What happens today is that companies have a series of manuals, notes, databases, documents where this type of information is stored. Whether it’s in Notion, an Office 365 folder, Google Drive, shared compliance documents or, in more structured companies, a database with all the company policies, the tools today are mainly DOCUMENTS.

So, what happens today is that access to current knowledge is given, with the invitation to read and digest it in some way.

If it’s videos or short clips within the onboarding process, they’re watched (as is the case with us, for example), but if I have policies or corporate guidelines in hundreds of pages and hundreds of documents (as is the case with our clients), it’s clear that no one is going to read them. It’s also because access is difficult and it’s impossible to remember everything.

What technology offers today, through tools that are increasingly accessible at enterprise level, is the ability for HR to create a knowledge base for all this information, put it on a SHARED DATABASE, make it more accessible, build an INTERACTION BOT on top of it (we’ve done this for clients and it works really well) and give people access by saying: “OK, I’ve trained you up to here, you should have accessed these hundreds of documents, and I accept that you didn’t because it’s impossible and almost nobody would do it. But if you have a question, I’m here to help you. Ask me a question, I’ll access the same knowledge base and I’ll give you the answer“.

We’ve seen that this is a really great evolution in the accessibility of information, and it SIMPLIFIES ONBOARDING, because it removes too heavy or perhaps unnecessary content in the early stages, restricts it to what’s essential, and shifts to 24/7 support everything that’s not needed in the very early stages and would also be disengaging.

AI can also help to create content and learning paths to bring everyone up to the same level, to be associated with an onboarding activity, and to animate them if you want to create informative videos.

In Digitiamo, we have short, in-depth content on “how we do it differently” for each of the main professional figures.

You can easily create content – both video and non-video – for study or direct interaction with a digital twin.

What technologies are used?

A combination of technologies is used: language models such as OpenAI, ChatGPT, Gemini, etc., together with tools for aggregating the knowledge base, which are semantic databases.

The combination of the two creates digital twin tools to which speaking faces can also be attached, with tools such as AIGen, for example.

This is the current set of technologies that allow us to do this work.

How do you see the development of the corporate DIGITAL TWIN? Are we still far from widespread use?

The Corporate Digital Twin is not yet very widespread, but it will be.

What we’re seeing is that once a company’s knowledge base is shared and accepted for certain profiles, it becomes easy – thanks to the AI tools available – to create a direct interaction with an INTERNAL VIRTUAL COACH to support the onboarding phase.

Today it’s very easy – starting from a job description, from what needs to be done within the role, maybe with line managers who have described the technologies and processes in more detail, and maybe with an ISO 9001 document that also provides additional details about the company’s processes – to feed all this information to an AI and automatically create a supporting digital twin for HR.

And every time I have a question on one of these topics, instead of going directly to HR, I would go to the company coach or the digital twin of HR.

What advice would you give to start implementing AI in onboarding processes? Where would you suggest to start?

For general employee support, apart from the need for specific training, you can start with a baseline of core values, compliance rules and basic company knowledge and evaluate the use of AI already available in company tools.

For the basics that every new employee needs to learn, the solutions tend to be close to the world of HCM (Human Capital Management), i.e., software that manages the entire relationship between the employee and the company, integrating AI capabilities to automate and improve HR management processes.

Before developing customised solutions, it’s best to check whether these AI functionalities are already available in the management software in use, in order to reduce costs and implementation time.

If these convenient AI features are not available, or if they aren’t sufficient, then it is necessary to contact SPECIALISED COMPANIES that produce AI tools or solutions that simply integrate the company’s knowledge and technologies to achieve the desired result, which is then delivered to the person.

This is the current way of working, while waiting for the technology to improve and mature, and for dedicated and vertical solutions to emerge for this type of activity, which are very supportive of the person and therefore very close to the HR world.

What are the most common mistakes in implementing AI tools at this basic level of onboarding? How can they be avoided?

The most common mistakes are:

Relying too much on AI without human oversight, which risks losing touch with individual needs.
Failing to personalise content, offering generic experiences that don’t address the real needs of the new hire.
Underestimating the training managers need to use AI tools, making it difficult to integrate them into existing processes

To avoid these mistakes, it’s important to combine technology and human interaction, gather continuous feedback and take an incremental approach to innovation.

How can AI be used to personalise onboarding and improve skills assessment?

In cases where specific training is required, there are two key moments:

The assessment phase, which we call post-assessment because it involves evaluations that were partly carried out during the recruitment phase.
The training and coaching phase.

The POST-ASSESMENT phase includes “what do I need, what is my role in the company, what do I need the person to do”. As this assessment should have been done during recruitment, I should already have some of this information.

I need to know what I want the resource to do, what the goal is, what tasks are in the daily activity list and what skills are required.

So, I need to make an assessment of the role and then fill any gaps accordingly.

The gap can usually be filled by TRAINING, COACHING or a HYBRID approach.

Thanks to AI, solutions are also being developed for both the assessment, learning path implementation and coaching.

Viblio for ASSESSMENT: you tell me where the person wants to go, I carry out an assessment, both soft and hard skills, and identify a path
LEARNING PATH IMPLEMENTATION: Various available LMS (Learning Management Systems) allow the creation of learning paths that are more or less dynamic, based on the individual and the content the company has
CONTENT CREATION: I may not have all the content I need for a specific training.

What AI tools can I use to create LEARNING TOOLS?

AI will increasingly enable the creation of hyper-personalised learning tools based on content.

So, if a person lacks a certain skill, I can create a hyper-personalised piece of content tailored to their needs using tools to create content – which I can review – and publish it so the next person has it available.

For example, I use BEAUTIFUL AI, which uses a prompt to create slides that can include voice recordings.

I can also create videos using tools like AIGEN, which takes content and allows you to create a video, including a face, to give a sense of interactivity.

There are also tools such as Google’s NOTEBOOK LM, which, when I post a topic, creates audio scenarios (similar to podcasts) that discuss a particular topic in a more interactive way, with two voices discussing a topic, allowing me to consume it in a more engaging way.

How can we ensure that the information delivered during an onboarding process is actually learned and applied in the long term?

Delivering content is always the more challenging aspect, because you need to VERIFY THE EFFECTIVENESS of what the person has received during onboarding.

There are AI tools that measure the long-term impact of what the person has actually learned within a given timeframe.

When I was at Docebo, we acquired a company that specialised in this type of metrics using AI. So, after a period of time, surveys were sent to people asking them what they remembered, and it did this using recognised patterns, backed up by psychological studies, to measure the actual impact of what was supposed to be transferred to the candidate and how much they retained.

What challenges do companies face when implementing an onboarding process based on current technology tools?

The main challenges are:

RESISTANCE TO CHANGE, especially from those used to traditional processes.
Difficulties in integrating with EXISTING HR SYSTEMS, especially in companies with outdated tools.
Risks related to PRIVACY and the management of SENSITIVE DATA.
Balancing AUTOMATION and HUMAN INTERACTION to avoid the onboarding process becoming an impersonal experience.

Overcoming these challenges requires a strategic approach and a strong focus on the USER EXPERIENCE.

Is there an aspect of onboarding where the potential contribution of AI is particularly underestimated, but will be critical in the coming years?

One of the most underestimated aspects is predictive analytics for long-term success.

AI could not only assess whether a person has the right skills for a role, but also predict their cultural fit, level of engagement and risk of turnover.

This would allow companies to proactively intervene to improve retention and integration of new resources.

What other EMERGING TRENDS in AI-powered onboarding do you see as particularly promising?

CORPORATE DIGITAL TWINS for more natural interaction between new hires and corporate systems.
PERSONALIZED LEARNING EXPERIENCES based on natural language processing and behavioural data.
IMMERSIVE ONBOARDING using AR/VR to simulate work environments and accelerate adaptation.
EMOTION ANALYSIS using AI to monitor stress and engagement levels during the first few weeks on the job.

Final question: “do you think onboarding will be an increasingly automated process in the future, or will the human factor remain central?”

Automation will increase, but the human factor will remain central.

AI will enhance the process, making it more efficient and personalised, but the human connection is irreplaceable when it comes to building a sense of belonging and motivation.

The future will be a balance between technology and relationships, where AI frees up time for more strategic and relational activities, without ever replacing the value of empathy and human interaction.

Ti è piaciuto il nostro articolo?

Condividilo con le persone a cui potrebbe interessare