The second part is about Application of Machine Learning, AI and Deep Learning in XM, Choose the most optimal model and AI/ML integration framework and Future plans and research and development directions

3. Application of Machine Learning, AI and Deep Learning in XM

Artificial intelligence (AI), machine learning (ML) and deep learning (Deep Learning) technologies are becoming increasingly popular Important levers to enhance XM experiences in the bank. AI/ML enables banks Analyze huge amounts of data about customer and employee behavior, from there find insight and Automate intelligent interactions, personalized service like never before. Below are the specific roles of AI/ML for each aspect of experience (CX, UX, EX, BX) and advanced AI techniques being applied. 

• AI enhances customer experience (CX): Artificial intelligence helps Personalize experiences at scale – something that manual methods cannot do. ML algorithms can analyze millions of transaction and interaction data to understand each customer's needs, from there Make suggestions for suitable products and services automatically. For example, AI chatbot and intelligent virtual assistants can provide Personalized financial advice, instant answers to customer questions anytime, anywhere . This enhances satisfaction and loyalty when customers feel understood and served promptly. AI also delivers 24/7 support service – Chatbot works continuously to help answer requests after hours, Ensuring customers are always served promptly without having to wait. In addition, ML is also used to predict customer behavior: for example, predicting the possibility of churn so that banks can proactively retain them with appropriate incentives, or forecasting credit demand to approach offers at the right time. Analyzing customer sentiment via voice calls or text responses using AI-powered NLP (natural language processing) also helps banks capture real-time satisfaction levels. Overall, AI/ML transforms CX management from reactive to reactive proactive, "understand before customers speak".
• AI helps improve user experience (UX): In UX design and optimization, AI is playing an increasingly larger role. First, AI helps Analyze user behavior on digital channels: ML tools can automatically collect click data, actions, heat maps... for detection Where users often encounter difficulties on the application/website. From there, the UX team knows where to improve (for example, a certain registration step that many users abandon -> needs to be simplified). Second, AI enables Flexible testing and interface customization: multivariate testing or AI algorithms can suggest themselves A/B testing plan and even Adjust the interface in real time based on user response. This creates UX Optimized for each customer segment. Additionally, AI delivers New UX features: for example Smart search (natural typing to find functions, powered by NLP), Automatic input suggestions, or application Vision AI to allow customers to open accounts with ID scanning and facial recognition instead of manual entry. It all helps Smoother, more seamless user experience. Another application is AI chatbot integrated right into the interface (like a guidance assistant) helps new users get acquainted with the application more easily, or quickly resolve questions during use - thereby reducing friction and increasing satisfaction when using digital products.
• AI drives employee experience (EX): For internal banks, AI and ML are also being leveraged to enhance the working environment and employee performance. One example is using an Internal virtual assistant: AI chatbot for employees, helping to quickly answer policies, procedures or professional instructions when needed, reducing time searching for information. AI also supports Personalized employee training and development – through analyzing performance data and career paths, the AI ​​system recommends suitable courses for each employee, or even plans promotions based on skills and interests. In everyday work, ML is possible Automate boring tasks (such as data entry, document reconciliation) thanks to RPA combined with AI, helping bank staff Reduce repetitive work load to focus on value creation tasks (customer care, financial consulting). More importantly, AI delivers Employee sentiment analysis tool through anonymous surveys or internal email tone analysis, helps management Catch early signs of dissatisfaction or reduce engagement to take improvement measures (for example, change working regimes, organize team-building...). Thanks to that, banks can proactively improve EX instead of just reacting after the employee quits. Even some pioneering banks have Draw the employee experience journey and use tracking technology at every touchpoint (onboarding, training, performance reviews, promotions, etc.), similar to how you do with the customer journey. In short, AI/ML is helping Personalize and optimize the experience for each employee, creating a more engaged, more efficient workforce – a prerequisite for delivering good CX.
• AI strengthens brand experience (BX): Brand management also benefits from AI, especially in the digital landscape where brand feedback data is abundant. First of all, AI enables Real-time brand reputation monitoring: social media listening tools using NLP and computer vision can scan through millions of posts and comments to Detect mentions about banking, analyzing tone of voice, positive or negative. Thanks to that, the brand team promptly grasps the media crisis or understands how customers are evaluating the bank's image. Second, AI helps Personalize marketing messages – an important part of BX. Instead of a generic message, AI can segment customers in detail and make suggestions Content suitable for each group (for example, the same loan ad but young customers will see different content than middle-aged customers). Approach Right person, right message, right time This makes the brand experience closer and more meaningful to customers. Third, AI supports assurance Brand consistency across multiple channels: smart digital content management systems can automatically check whether documents and advertisements comply with brand standards (colors, logos, tone of voice), or even suggest edits to synchronize images. Finally, AI also provides Predictive analysis of brand value – for example, using a model to predict that if the NPS score increases by X, the brand awareness index can increase by Y, helping leaders guide brand strategies based on data. All of the above applications aim at one goal: Build a trustworthy, customer-understanding and consistent banking brand, thereby improving BX. In fact, the conversational AI model (chatbot) not only for CX but also plays a role in BX - when the chatbot represents the brand chatting with customers, if well trained, it will create a positive brand impression (modern, useful). Recent research also suggests Conversational AI models have the potential to positively impact customer and brand experiences, but further research is needed to clearly understand the influencing factors (Providing a Model of Customer Experience Management Based.pdf).

Deep Learning and advanced methods: Deep learning, with its power to process unstructured data and learn complex patterns, is opening up possibilities for groundbreaking XM application in the bank. Multi-layer neural networks are capable of detecting hidden patterns in customer behavior that traditional methods miss. For example, model RNN (Recurrent Neural Network) sâu has been researched to Analyze the chain of events in the customer journey, helps banks clearly understand key touchpoints and predict your guest's next move in a complicated journey. Deep learning cũng Upgrade NLP capacity: modern language models (such as BERT, GPT) allow chatbots to understand customer intentions more accurately, or analyze emotions in millions of text responses with high accuracy. In terms of computer vision, deep learning helps Face and handwriting recognition... application in eKYC, both increasing security and smoothness (customers open an online account in a few minutes). Special Generative AI (generative AI) is expected to "transform" the banking experience. Generative models like GPT-4 can interact naturally like humans, providing a Smart virtual financial assistant for each customer. For example, Gen AI can act as a “personal financial advisor” Answer all questions and proactively make suggestions based on customer data. This use case Innovation has appeared: banks use Gen AI to Agent Assist by summarizing information and suggesting real-time feedback; Analyze a series of responses to Classify and prioritize complaint resolution (Quality Automation); Super personalization experience - deeply understand preferences and behaviors to completely customize products and services for each individual. All on target Maximize simplification and personalization of the customer journey at every touch point. The expected result is Outstanding CX and stronger customer engagement. In fact, experts say “Gen AI in banking will elevate personalized service and deliver superior CX, at the same time pure Simplify operations and speed up request processing”. Of course, along with that come challenges that need to be managed (e.g. bias, security, transparency of AI models)..

In short, AI/ML/DL is being strongly applied on all fronts of banking. From work, analyzing experience data (feedback, behavior) in a smarter way, interactive automation (chatbot, virtual assistant) arrives trend forecast (early detection of needs, dissatisfaction points) – AI helps banks Deep understanding of customers and employees to make quick decisions and enhance the experience. New generation AI and deep learning methods will continue expanding XM's borders, making ideas like “personalized banking for a single customer” a reality. The role of AI in XM can be summarized: transforming experience management from an emotional art into a proactive, data-driven and predictive science.

4. Choose the most optimal model and AI/ML integration framework

Based on the above analysis, it can be seen All CX, UX, EX, BX aspects are important for the banking industry. However, to determine The most optimal XM model Need immediate priority, we consider according to the following criteria: affects business goals, scope of impact, feasibility and implementation efficiency. According to these criteria, customer experience management (CX) emerging is The focus model should be given top priority in banking:

• Direct impact on business performance: CX is associated with satisfaction, loyalty and revenue from customers - the bank's vital goals. Enhance CX will Instantly increase loyalty and reduce churn, Quick impact on business results. Meanwhile, UX, EX, BX are important, but the impact is often more indirect or long-term (for example, good BX improves brand awareness but takes time to convert into revenue).
• Comprehensive and comprehensive scope: CX has a wide scope, including both UX elements (digital channel experience) and human service elements (influenced by EX). Investing in CX enables At the same time, it touches many aspects: If you want good CX, you must improve both UX and employee interaction. Therefore, focusing on CX will naturally lead to upgrading UX and EX.
• Feasibility and measurement: CX management already exists in many maturity measurement frameworks (such as NPS, CES, CSAT) and support tools, so banks can easily establish continuous improvement programs. CX is also easy to link with available customer data to apply AI analysis. On the contrary, EX or BX are more difficult to measure (eg, measuring employee emotions, brand value), so it is difficult to deploy quickly and synchronously in the early stages.
• Proven effectiveness: Many studies and practical cases confirm the benefits of focusing on CX High ROI for the bank. As stated, 82% of banking sector businesses consider CX a top priority to retain customers. The leading banks in the market (Capital One, DBS, OCBC...) all built strategies around customers and achieved impressive growth. Even banks have seen improved customer experience and branding go hand in hand outstanding revenue (3.5x) compared to competitors. This shows that investing in CX is an effective choice.
• Suitable for the current context: Currently, when digital transformation and competition is happening strongly, customer experience is the hottest front. Customers have too many financial service options, so Whoever delivers the better experience wins. Therefore, focusing on optimizing CX is a strategy that matches market trends.

However you choose CX is the priority focus, it should be emphasized that UX, EX, BX aspects are not neglected. Instead, banks should take a directional approach of integration - take CX as the center and use UX, EX, BX as supporting "levers".. The end goal is Synchronize all banking-related experiences so that customers and employees are satisfied and the brand is strengthened.

CX model integrating AI/ML – Proposed framework

To realize optimal CX management, banks need a framework that integrates AI/ML into CX programs. Below is a step-by-step suggestion for implementing this model:

1. Create a CX vision and set target metrics: First, the bank needs to determine the ideal customer experience you want to achieve and tie it to specific goals (e.g. increase NPS by X, reduce complaints by Y, etc.). Set up Key CX indicators (KPI) like CSAT, NPS, CES, churn rate... as a measure of success. At the same time, formed a CX dedicated team (Chief Experience Officer, customer experience department) to lead the initiative. The CX vision should be clearly communicated throughout the organization so that all departments work towards the same goal.
2. Collect omnichannel experience data and build a unified data platform: At this step, the bank deploys systems to Collect CX data continuously from every touch point: post-transaction customer surveys, app/website feedback, customer service call data, chatbot interactions, social media feedback, complaints, as well as employee experience data (internal survey) and brand data (market information, social networks). All data is put into one centralized data platform (data lake or CX platform) to get a 360 view of the experience. This integration is important for breaking down silos – for example, connecting UX data (like app usage behavior) with CX feedback data for cross-analysis. This is also the time to apply Experience Management software tools (e.g. Qualtrics XM, Medallia) to support collection and management of experience data.
3. Analysis and Insights using AI/ML: With abundant and diverse data, the next step is using machine learning to analyze and discover insights. Specifically:
   • Apply AI analyzes emotions and classifies responses: use NLP to process thousands of customer comments, automatically classify complaint/praise topics, measure positive/negative ratios over time.
   • Applied ML predicts behavior: build models (like decision tree, random forest) to predict the churn risk of each customer, predict which customers are likely to buy which product next,... As research has shown, algorithms such as decision tree Very effective and easy to understand. Classify customer perceptions and predict experience results (Providing a Model of Customer Experience Management Based.pdf).
   • Use AI to find the root cause (root cause analysis): for example, AI detects that the common point of transactions with low CSAT is that customers have to wait >5 minutes. From there, determine specific blockage in the journey.
   • Build smart dashboard: Integrate AI into the CX management dashboard to automatically warn when an index experiences a sudden decrease, or when there is a new trend. For example, AI can warn that "complaints about mobile applications increased by 30% this week, mainly related to login errors". These proactive analytics and alerts help banks Understand CX issues instantly and have a data-based decision-making basis instead of sentiment.
4. Action to improve experience - applying AI to real-life interactions: Based on the insight from step 3, the bank proceeds specific CX improvement actions, while integrating AI into the execution process:
   • Personalize service: Deploy AI models recommend products/services suitable for each customer in daily transactions (for example: suggesting to open a savings book when you see a customer has spare money). Like ING's case, the application AI recommendation engine Analyze transaction and behavioral data to suggest products has helped increase customer interaction and satisfaction.
   • Optimize important touch points: If the analysis shows that the online credit card registration step has low CSAT, proceed UX redesign of that step (reduce the number of information fields, support automatic filling). Incorporate A/B testing and adjust based on real-time feedback (can use AI to automatically choose the best version).
   • Enhance EX at service point: For example, if a customer complains that the teller's advice is not good, the bank can provide additional training for relevant staff. Leverage AI to Simulate service situations (Through e-learning, there is a chatbot that plays the role of a difficult customer for employees to practice). Simultaneously, empower employees frontline with information: equipped with internal mobile applications providing 360 information about customers (thanks to AI synthesized from CRM data), helping employees personalize on-site service.
   • Adjust your brand message: If data shows that customers misunderstand service fees, the marketing department may need to communicate more clearly (simple infographic example). Or if a trend is discovered that customers are interested in "green banking", the brand can proactively bring messages about green initiatives into the experience (website/app adds an ESG section...). These moves ensure BX is consistent with customer expectations, increasing sympathy.
   • Support and process automation: Wider deployment of AI chatbots for customer support on website, Facebook, mobile app... to quickly answer frequently asked questions, resolve basic requests (balance lookup, card blocking...). This fits improved experience (fast, 24/7) just reduced the operator load. For internal processes related to customers (loan approval, account opening), integrate AI RPA to increase processing speed, reducing waiting time for customers.
5. Re-measure and continuously improve (Learning Loop): After each deployment action, need Measure the results that impact the experience through established CX metrics. Use ML to compare before-after performance (for example, after improving the card registration UX, did the CES – effort score for this step increase?). Collect new customer feedback on changes. Since repeat the process: update the AI ​​model with new data, fine-tuning the solution. This is a continuous loop: Measure → AI Analysis → Action → Measure again, bank guarantee continuously learn and adapt to enhance CX. An important point is that the feedback flow must be fast and flexible – culturally “Test fast, fail fast, learn fast” should be encouraged.
6. Expanding the range of integrated XM: When the CX + AI framework operates smoothly and produces results, banks can expand to other aspects of experience creating a comprehensive XM program. For example, similar implementations for EX (collect employee feedback periodically, use AI analysis to improve HR policies), and BX (using market data to measure brand recognition, associated with CX strategy). The goal is to eventually establish one unified XM ecosystem, where Every customer and employee experience is continuously measured and improved with data.

The above framework requires a thorough investment in technology (data infrastructure, AI) and a change in management thinking (customer-centricity and data-based decision making). However, when done right, it helps the bank Optimize customer experience systematically and proactively, thereby achieving both short-term effectiveness (satisfaction, revenue) and building long-term competitiveness (outstanding experience, strong brand). In the context that the world's leading banks have been moving in this direction, early adoption of the model CX integrated AI will be the optimal choice for breakthrough for banks.

5. Future plans and research and development directions

After determining the key XM model (CX) and implementation framework, the next stage of research and application will include some main directions as follows:

• Develop research into scientific articles: From this detailed research foundation, it is possible to proceed systematized to become a complete scientific study. Specifically, it is necessary to build a clear methodology (for example, comparative analysis framework of XM models, proposed AI integration model), collect empirical data from banks (through expert surveys, case study analysis as stated by ING and Origin Bank) to Verify the arguments. The research results can contribute to the academic literature Experience management in banking and finance, especially emphasizing the role of AI/ML in the new era. Jobs published in specialized journals (on management information systems, financial services marketing) will be considered. One specific direction is construction quantitative model Test the relationship between AI adoption and customer experience quality (for example, measuring NPS increase when AI chatbots are available, or the impact of EX on CX in the Vietnamese banking industry). Gaps like interaction between AI and customer experience There is still little research (Providing a Model of Customer Experience Management Based.pdf), so this is an opportunity to make an academic contribution.
   
• Next steps for implementation XM in bank: From a practical perspective, after the planning and framework design stage, banks need a specific implementation roadmap. Important steps include:
   
  • Controlled Pilot: Choose a customer segment or a few branches to test an AI-integrated CX program on a small scale. For example, pilot an AI chatbot and feedback collection system in 5 urban branches for 3 months, measuring the impact.
  • Evaluate and edit: Based on pilot results (CX indicators, participating employee feedback), refine the process and AI model to suit the actual context of the bank. Ensure issues are fixed before scaling (e.g. chatbots need additional expertise? employees need more training on CX culture?).
  • Scaling up: Official deployment throughout the system based on pilot experience. At the same time, invest in upgrading IT infrastructure if necessary (data warehouse, more powerful analytics tools) to process data from a larger network. This stage is necessary mass training Inform employees about the new XM philosophy, ensuring everyone understands the goals and how it operates (for example: employees proactively collect customer feedback, how to use the AI ​​dashboard when serving...).
  • Continuously monitor and refine: Once deployed, maintain the mechanism Regular meetings about the experience (monthly CX review) at all levels of management to monitor progress and problems. Always update AI algorithms according to new feedback (continuous learning model). Job Maintain a culture of continuous improvement is the key for XM to take root.
• In terms of project management, the method can be applied agile for small initiatives within the XM framework (related to technology) to quickly produce results and adapt to changes.
   
• Deeper research on technological factors and application expansion strategies: The field of XM incorporating AI in banking is still evolving, so it is necessary to continue researching new technology trends and ways to expand strategies. Some in-depth research directions include:
   
  • Advanced AI applications: Research the possibility of applying new technologies such as Generative AI in banking services. For example, use GPT-4 to create virtual financial advisor Highly personalized, or create real-time interactive marketing content for each customer. Assess their impact on the experience (will customers be comfortable with fully AI advice? where is the line between automated and human?).
  • Emotional artificial intelligence: Learn about integration AI recognizes emotions (from voice and face during video calls) into the VIP customer service process, helping staff immediately grasp customer emotions to adjust their approach.
  • New technology in EX: Usage survey metaverse or virtual reality (VR) in banking staff training – does role-playing experience improve skills and engagement more than traditional training? Or use wearable AI to monitor employee health (especially sedentary office employees) as part of EX care.
  • Comprehensive digital transformation strategy towards XM: Research successful cases around the world such as BBVA, Bank of America – How have they changed their organizational structure and processes so that XM is instilled in their business strategy? From there, lessons can be drawn for other banks.
  • Risk management and AI ethics in XM: When expanding AI in customer interactions, research is needed Code of ethics, data security, privacy. For example, how far is it reasonable to use personal customer data for personalized AI? How to solve when AI makes wrong decisions that negatively affect the experience? Building an AI governance framework in the financial sector aims to ensure AI serves well but does not violate regulations or ethics.

Finally, an important direction is Long-term measurement of XM effectiveness: research the correlation between experience indicators (such as NPS, eNPS for employees, BX index) with financial efficiency of banks (revenue growth, market share, brand value) through econometric models. This helps reinforce quantitative case studies that XM investments deliver tangible results, convincing stakeholders to continue supporting this strategy.

In short, the next journey requires everything Practical action and academic research in parallel. In practice, piloting and replicating the XM model in banks with the support of AI/ML will create the foundation for a Experience-oriented banking. Regarding academia, expanding research into new aspects of XM (especially the role of AI) will contribute knowledge to the industry and support better practice. With the current development of technology and customer expectations, constantly researching and improving XM will help banks Leading trends, creating outstanding experiences and sustainable success in the future.

Typical references:

• Rahmani, A. et al. (2024). Providing a Model of Customer Experience Management Based on Knowledge Management Models in Fintech Using Machine Learning – Journal of Systems Thinking in Practice (Providing a Model of Customer Experience Management Based.pdf) (Providing a Model of Customer Experience Management Based.pdf). (Customer experience management model based on knowledge management and ML in fintech).
• Forrester Research (2025). Brand And Customer Experience Together Power Growth – Forrester Report
• Maze (2023). Banking Customer Experience Trends To Watch in 2025 
• Qualtrics XM Institute (2024). The State of CX in Retail Banking 
• Blink UX (2021). Why Good UX Is Essential for Banks 
• Believe in Banking (2023). Banks Are Building Better Employee Experiences  
• UXDA (2024). Restoring Trust in Banking: Digital Experience Branding 
• Nexgen Banking (2023). The Role of AI in Enhancing Customer Experience in Banking
• WNS (2023). Harnessing Generative AI in Banking to Transform Customer Experience
• DigitalDefynd (2025). AI in Banking – 20 Case Studies