AI APPLICATIONS IN SIX SIGMA: INTEGRATING MACHINE LEARNING INTO DMAIC FOR TELECOMMUNICATIONS PROCESS IMPROVEMENT

Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [340] AI APPLICATIONS IN SIX SIGMA: INTEGRATING MACHINE LEARNING INTO DMAIC FOR TELECOMMUNICATIONS PROCESS IMPROVEMENT Vamsi M Nellutla, Dallas Data Science Academy, Irving, Texas 1. INTRODUCTION Six Sigma has become one of the most powerful data-driven process improvement methodologies and has provided organizations with a structured method on how to reduce variation and enhance quality. The DMAIC cycle of Define, Measure, Analyze, Improve and Control has long been a potent source of systematic problem solving and operational excellence as Bukhari and Akhtar (2024) explain. However, the contemporary industrial and service environment has greatly transformed, and the traditional Six Sigma tools are growing to be questioned by the magnitude, velocity, and intricacy of the current data environments. The transition is particularly evident in the industries that have high-volume and high-velocity data in their operations, like the telecommunications. Channappagoudar, Bhat and Gijo (2025) claim that telecom systems are producing continuous data streams of network gear, customer interactions and digital interfaces, which puts an analytical load much larger than that of the original design of the Six Sigma techniques. Industry 4.0 technologies, as these authors observe, are changing the quality improvement expectations, and organizations have to incorporate intelligent, automated, and predictive into their DMAIC processes. Although the actual change is still in the realm of Artificial Intelligence (AI) and Machine Learning (ML), these two elements are gaining popularity as the facilitators of this process. According to Kabeer (2025), AI, specifically machine learning, can be used to assist in continuous process improvement as it can help find patterns and process anomalies that could otherwise be challenging to identify. On the same note, Sood and Dhull (2024) affirm that AI compliments the Six Sigma in bolstering predictive analytics, therefore, improving the ability of DMAIC to foresee defects, variability and performance concerns before they become aggravated. The concept of AI inclusion in Six Sigma belongs to an active trend of the predictive, digital quality management. Santacruz (2023) emphasizes the fact that integration of ML with the Six Sigma helps organizations to transition to zero-defect performance by changing problem solving to simulation and proactive forecasting. Similarly, Maged, Haridy, Awad and Shamsuzzaman (2023) present the actual application of machine-learning-supported Six Sigma tools, indicating that AI improves the flexibility and analysis depth of the DMAIC cycle in various working scenarios. Those changes can be characterized by the fact that Escobar, Macias and McGovern (2022) describe the process of turning Six Sigma into Quality 4.0 when (1) intelligent automation, (2) advanced analytics and digital technologies will become more enriching each phase of the DMAIC process. Similarly, Mansour, Abohashima, Elkhouly and Scow and others (2025) demonstrate that the combination of AI-based defect prediction systems and Six Sigma promotes the quality control in real-time and shortens the improvement cycles considerably. Kabeer (2025) goes on to say that the use of AI to find anomalies and simulate processes results in DMAIC being more efficient and responsive, particularly in dynamic environments. One of the areas where such innovations are likely to be useful is in the telecommunications. As illustrated by Uluskan and Karski (2022), machine-learning-enhanced DMAIC, commonly known as Predictive Six Sigma, assists telecom operators to detect the reduction of the services in the early stages, forecast equipment malfunction and network optimization. As networks grow more complicated with the growth of 4G, 5G and IoT networks, the power of AI to process huge amounts of operational data turns it into a logical extension of the organised approach of Six Sigma. Generally, there is a definite overlap in the literature: AI reinforces Six Sigma by enhancing the predictiveness of the DMAIC cycle, streamlining and adapting it to the reality of modern telecommunications data. When telecom operators are in need of enhancing reliability of their service provision, minimizing downtime and improving customer experience, a promising direction to the digitally enabled, ongoing process improvement is the incorporation of machine learning in DMAIC. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [341] 2. INTRODUCTION TO SIX SIGMA AND THE DMAIC FRAMEWORK. Six Sigma is constructed around the idea that variation should be minimized and consistency within organizational procedures enhanced by relying on the data-driven decision-making process. According to Bukhari and Akhtar (2024), the approach is based on the principles of statistical thinking, methodic problem solving, and the need to increase customer satisfaction. It is strong due to its systematic way of detecting inefficiencies, the measurement of performance, the examination of the causes of defects, the actualization of improvement, the putting in place the long-term control mechanisms. 2.1 Six Sigma Foundations Removal of variation in processes to ensure the perfection of performance almost to perfection is the main aim of Six Sigma. Organizations use a rigorous system of knowing and enhancing processes by reaching a goal of 3.4 defects per million opportunities. To Channappagoudar, Bhat and Gijo (2025), the following tools have always been used to support these principles: process mapping, control charts, Pareto analysis and hypothesis testing, which assist organizations in making informed decisions that are supported by statistical evidence. Maged, Haridy, Awad and Shamsuzzaman (2023) also demonstrate that the formal quality of Six Sigma enables teams to traverse tricky operational issues in manufacturing and service settings. 2.2 The DMAIC Methodology The main cycle of the implementation of Six Sigma is represented by the DMAIC framework, which includes the following steps: Define, Measure, Analyze, Improve and Control. Define: According to Sood and Dhull (2024), this stage is aimed at comprehending the issue, defining the needs of the customers, and setting the project objectives. This can be applied particularly to the context of telecom where issues of service quality which include call-drop rates or network congestion is framed. Measure: This step will entail gathering baseline data and quantifying the issue. Kabeer (2025) states that the traditional Measure tools consist of the process capability analysis, measurement system analysis, and the development of detailed data maps. Analyze: The aim of this is to identify the underlying causes of the variation. According to Escobar, Macias and McGovern (2022), classical Analyze methods, including regression analysis and fishbone diagrams, are formal methods of determining the drivers of defects. Improve: Solutions are created and put into practice once the causes are known. According to Santacruz (2023), traditional improvement activities are based on controlled experimentation, design of experiments, and specific redesigns of processes aimed at the reduction of variation and the optimal performance. Control: Last but not least, organizations maintain improvements using monitoring systems, control charts and standard work procedures. As Mansour, Abohashima, Elkhouly and others (2025) demonstrate, effective Control will guarantee long-term stability and avoid a decline of performance levels to the initial ones. These phases allow organizations to come up with complex problems in a logical and repeatable way; that is, these stages make up a complete cycle. 2.3 Traditional shortcomings of Six Sigma in Telecom. Although the Six Sigma has been very effective in several industries, its classical form has been challenged where dynamic and high frequency data is being generated like in the telecommunication field. As Uluskan and Karşi (2022) note, telecom networks generate gigantic amounts of data, such as call-detail records, real-time network logs, and other types, which are often beyond the analytical power of many traditional Six Sigma instruments. An illustration of this is that manual data collection and data analysis processes find it difficult to identify latent anomalies or intermittent failures in systems with rapid change. Equally, Kabeer (2025) states that human analysis analysis is likely to be sluggish and erratic especially when handling non-linear or high-dimensional data trends. The nature of telecom processes also changes rapidly in response to the changing demand in the traffic, weather, the mobility of the devices and the changing behaviour of the customers hence the application of the fixed analytical tools is not sufficient. According to Escobar, Macias and McGovern (2022), such complexities demonstrate the inability of the purely statistical methods and emphasize the necessity of the AI-based solution that is capable of processing the data at the scale and provides predictive insight and automates the analysis. According to Bukhari and Akhtar (2024), a lack of real-time intelligence may decelerate a DMAIC cycle and minimize the quality of root-cause detection. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [342] Briefly, the increasing data volume of telecommunications reveals vulnerabilities in conventional Six Sigma frameworks, which is why there is a strong rationale to use AI and machine learning to develop a complex level of analytical proficiency, speed, and accuracy. 3. The role of Artificial Intelligence and Machine Learning in Quality Improvement. AI and ML have taken over as the main point of modern quality improvement values, especially in industries with high data volumes, like telecommunications. The transformation towards quality management that is based on digital tools necessitates, as Channappagoudar, Bhat and Gijo (2025) explain, the use of powerful analytical tools, which can empower organizations to obtain insights of large, rapidly evolving datasets. The classical implementation of Six Sigma, though an effective approach, was not initially created to work with the magnitude and the complexity of modern operational data. This gap is filled by providing AI with computational intelligence, predictive functionality and automation to the quality management process. 3.1 Principles of AI/ML Relevant to Six Sigma. AI is a field of study that involves a variety of computational approaches, which allow machines to learn information and make informed decisions. One of its most influential subfields, machine learning, enables devices to recognize patterns automatically, recognize events and forecast their conclusions without being explicitly programmed. Kabeer (2025) points out that supervised learning, clustering, neural networks and anomaly detection are the new ML tools that could be used to analyze operational inefficiencies. In the same vein, Sood and Dhull (2024) point out that deep learning structures can improve the capability to process the complex, nonlinear data that were previously challenging to understand by using Six Sigma tools. The other applicable AI method is reinforcement learning. Reinforcement learning enables systems to modify their behavior over time based on learning, which is why the reinforcement learning is an optimal choice in continuous optimization problems typical in telecommunications, as described by Escobar, Macias and McGovern (2022). These analytical techniques broaden the investigation capabilities of DMAIC because they permit systems to identify delicate trends, dynamic interrelations and previously neglected variables of variation. 3.2 Why AI Enhances Six Sigma The application of AI to Six Sigma is informed by the fact that it handles large, heterogeneous and high frequency data in a manner that is beyond the analytical capacity of human beings. According to Bukhari and Akhtar (2024), AI can be viewed as a complement to DMAIC because it enhances the accuracy of root-cause analysis and allows predictive quality control. This capability to predict defects prior to their happens to be a major development over the traditional Six Sigma which is mainly reactive. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [343] Maged, Haridy, Awad and Shamsuzzaman (2023) confirm that models based on AI are better than the manual ones at analyzing complex manufacturing and service data, providing more profound insights into the diagnosis. Simultaneously, Mansour, Abohashima and Elkhouly (2025) demonstrate that AI-based prediction defect systems can support the Improve and Control phases with more stable, automatic and real-time monitoring of the processes. These capabilities combined enable DMAIC to move beyond the realm of analysis in stature to active intelligent quality improvement. 3.3. Telecommunications AI Uses. One of the spheres in which AI provides the greatest quality improvement gains is telecommunications. According to Uluskan and Karsi (2022), telecom networks produce massive amounts of operational information in call-detail records, network equipment, and customer interactions and IoT devices. Machine-learning applications are particularly useful to analyze these massive datasets to detect defects, forecast service degradation and run tasks more efficiently. According to Kabeer (2025), AI-based methods allow telecom operators to find anomalies like the sudden increase in traffic, the appearance of the network congestion or maladaptive behavior of equipment. Santacruz (2023) further explains that predictive models are able to foresee network failures even before they take place and thus minimizing the downtime as well as enhancing service reliability. Regarding the customers, Sood and Dhull (2024) mention that AI-enhanced sentiment analysis can assist organizations in comprehending their customers needs more precisely, which allows them to better provide services to their customers in the way they expected it. On the whole, it is evident in the literature that AI is introducing high-quality analytical depth, velocity, and precision to Six Sigma practices, particularly in sectors such as telecommunications, where quality enhancement is becoming a more and more reliant aspect of real-time knowledge and foretelling intelligence. 4. THE IMPLEMENTATION OF AI WITHIN THE DMAIC PROCESS. The addition of Artificial intelligence to the DMAIC model will be a major shift in the Six Sigma practice, as it will allow companies, particularly telecom operators, to shift their quality improvement efforts towards being proactive and automatic. Bukhari and Akhtar (2024) state that the integration of AI into all the DMAIC phases increases the accuracy of the analysis, decreases the time per cycle, and makes the process optimization less biased. This paragraph will examine how the application of AI technique enhances each step of the DMAIC process. 4.1 AI in the Define Phase Define phase establishes the basis of a Six Sigma project by defining the problems, their customer requirements as well as outlining the project goals. AI also promotes this step by making it more accurate and thorough in identifying the problem. Sood and Dhull (2024) note that organizations are increasingly using Natural Language Processing (NLP) to process customer feedbacks, transcribed texts in call-centres and online complaints. NLP devices are able to bring out sentiment, classify issues and discover repeat service issues that would otherwise remain hidden under the conventional process. This voice-of-customer analysis using AI to help telecom companies prioritize areas of improvement, as Channappagoudar, Bhat and Gijo (2025) explain, is based on the real demand and customer expectations. In addition, Kabeer (2025) notes that machine-learning classification algorithms can automatically categorize the types of problems, extract issues with potential high impact and assist project scopes. This minimizes the subjectivity that is normally denoted when defining telecom service problems like network congestion, call-drops or service lags. 4.2 AI in the Measure Phase The Measure phase is the stage that is traditionally aimed at gathering credible data and defining the baseline performance of a process. The application of AI makes this stage much more powerful, as it allows the automated, scalable and real-time data capture. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [344] Telecommunications networks produce streams of enormous data streams of sensors, towers and OSS/BSS systems. According to Uluskan and Karşı (2022), AI-based data pipelines enable organizations to capture and work on these large volumes with the least human involvement. Machine learning is also important during data cleaning, such as noise reduction algorithms and anomaly detection algorithms can be applied to the network data to remove noise and inconsistencies before the analysis process starts. Maged, Haridy, Awad and Shamsuzzaman (2023) demonstrate that the precision of downstream analysis is enhanced by using more accurate feature extraction through the assistance of ML-based measurement systems. In the meantime, Mansour, Abohashima and Elkhouly (2025) show that AI-driven dashboards can show live network performance measurements, providing telecom teams with an insight into the service variations in real-time. 4.3 AI in the Analyze Phase The step of Analyze is when the causes of variation or defects are determined, which is where AI has significant benefits. The nonlinear data patterns that present themselves in telecom systems typically do not lend themselves to traditional Analyze tools like regression, Pareto charts, etc. According to Escobar, Macias and McGovern (2022), machine-learning models have a higher diagnostic power, as they reveal unknown associations between data in high-dimensional space. The predictive analytics can determine the causes of dropped calls, customer churn, network outages or traffic bottlenecks with a lot more precision than a manual one can. Kabeer (2025) further notes that clustering algorithms can be used to categorize telecom faults into valuable groups, which can be targeted to effect a specific intervention. Network logs can be analyzed using deep learning models to identify early signs of failure and causal inference models can be used to distinguish between correlation and causation. Santacruz (2023) also remarks that AI-based simulations enable organizations to simulate the effect of certain variables on the performance of services in advance prior to implementing changes. 4.4 AI in the Improve Phase During the Improve stage, the focus is on creating and testing solutions which would maximize the performance of processes. Predictive modeling, simulation and automation enable AI to expedite this stage. The so-called reinforcement learning, as explained by Escobar, Macias and McGovern (2022), can test thousands of possible improvement strategies and find the most effective one automatically. This in telecommunications may be the optimization of the signal routing, or balancing of the network loads or better resource allocation. Complex decision-making in fields like frequency management, tower-tower handoffs and bandwidth allocation can be achieved with the use of genetic algorithms and optimization engines that are highlighted by Kabeer (2025). According to Santacruz (2023), digital twins are artificial models of telecom networks based on AI to allow teams to test the ideas of improvement without affecting real-world operations. 4.5 AI in the Control Phase The Control phase makes sure that the changes are maintained in the long term making sure that processes will not go back to their initial position. This phase is enhanced greatly by AI by constant observation and anticipatory management. Mansour, Abohashima and Elkhouly (2025) demonstrate that when comparing ML-driven control charts with traditional statistical control tools, the former will indicate any performance deviation sooner than the latter before the defects get out of control. Drift detection algorithms are used to ensure that the model is upheld as the behavior of customers or the condition of the network varies. On the same note, Bukhari and Akhtar (2024) indicate that AI-based monitoring systems facilitate real-time monitoring of telecom operations so that the gains of improvement can be retained. Also, Uluskan and Karski (2022) include that predictive maintenance models during the Control phase avoid the occurrence of equipment failure, which minimizes downtime and ensures the quality of the service. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [345] Table: AI Integration Across the DMAIC Phases DMAIC Phase Traditional Six Sigma Activities AI / Machine Learning Enhancements Benefits for Telecommunications Define Identify problems, customer needs, project goals NLP for Voice of Customer analysis; automatic issue classification; sentiment analysis More accurate identification of customer pain points; faster problem clarification Measure Collect baseline data; assess measurement systems; map current performance Automated data capture from OSS/BSS; ML-based data cleaning; anomaly detection; real-time dashboards Higher data accuracy; real-time visibility; scalable measurement of massive network datasets Analyze Root-cause analysis using statistical tools; identify sources of variation Predictive modelling; clustering for fault segmentation; deeplearning pattern detection; causal inference algorithms More precise diagnosis of network faults; early detection of congestion and outages Improve Develop solutions; test changes; optimize processes Reinforcement learning; genetic algorithms; digital twins; AIbased optimization engines Intelligent resource allocation; simulation-driven improvements; automated network configuration Control Monitor performance; use control charts; standardize procedures ML-driven control charts; drift detection; real-time alerts; automated performance monitoring Sustained improvements; proactive issue prevention; stable network quality 5. TELECOMMUNICATIONS APPLICATIONS Telecommunications business is a very dynamic and data-driven field and so one of the most appropriate fields of applying Artificial Intelligence to the Six Sigma. Operational, behavioural and performance-related data are produced in masses per second by the telecommunications networks. Creating actionable insights with the help of these datasets is important as it allows maintaining the reliability of the network and the quality of the offered services (Uluskan and Karşi, 2022). AI based DMAIC can help telecom operators move away towards reactive troubleshooting into predictive and prevention based quality control. In this segment of the paper, the author discusses the major use cases where AI and Six Sigma have been most effective in their integration. 5.1 Predictive Quality Control Telecom networks undergo constant changes in the load of traffic, signal quality, environmental factors and hardware capabilities. Conventional methods of quality-control are frequently not able to identify developing issues at an early enough stage to avoid degradation in service delivery. Santacruz (2023) notes that machine learning-based predictive models are capable of detecting indicators of equipment breakdown, network overload or signal degradation at extremely early stages when the conditions are not yet critical. According to Maged, Haridy, Awad and Shamsuzzaman (2023), deep learning algorithms operating based on network logs and tower sensor data are able to identify very small anomalies that have the potential to trigger outages. Mansour, Abohashima and Elkhouly (2025) also demonstrate that AI-based prediction of defects in realtime is a massive enhancement of the Improve and Control phases due to the possibility of implementing early intervention. Practically, this will assist telecom operators to minimize and avoid dropped calls, minimize downtime and ensure and maintain quality of service (QoS). 5.2 Improving Customer-Centric Processes. The telecommunications industry is a performance area that is very critical based on customer satisfaction, which depends on the reliability of the services, response time and the general user experience. The issue of AI-powered DMAIC has a significant part in enhancing customer-facing processes. Sood and Dhull (2024) emphasize that the Natural Language Processing augments the Define and Analyze stage as it allows telecoms to understand the customer complaints, social-media interactions and call-centre calls transcripts on a mass-scale. This automatic voice of customer analysis shows the trends in user dissatisfaction which might not be realized with the help of traditional manual analysis. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [346] Kabeer (2025) also includes the fact that machine-learning models can be used in intelligent call routing and workload predictions in customer service. Such systems cut down the average handling times and increase firstcall resolution by making sure that the customers are routed to the most appropriate agents or automated solutions. Bukhari and Akhtar (2024) state that to a certain extent, predictive analytics are capable of foreseeing customer dissatisfaction on the verge of the problem emerging, enabling the company to take a proactive measure. 5.3 Network Optimization The optimization of the performance and efficiency of telecommunications networks is one of the most effective uses of AI-integrated DMAIC. The loads of traffic differ by place and time, and dynamic responses are necessary, which cannot be achieved by traditional tools of Six Sigma. According to Escobar, Macias and McGovern (2022), network routing can dynamically change reinforcement learning and optimization algorithms to alleviate congestion. According to Kabeer (2025), AI-based loadbalancing models provide fair allocation of network resources among towers and particularly in the high demand regions. In the same manner, Channappagoudar, Bhat and Gijo (2025) show that Industry 4.0 solutions, including IoT sensors and AI-based monitoring services, assist telecom operators with detecting bandwidth limitations and actively adjusting the parameters to ensure service quality. According to Santacruz (2023), digital twins recreate the impact of control variations, making it possible to experiment without interfering with live networks. 5.4 Telecom Case Examples Despite the fact that exact implementations differ according to the operator, current studies indicate that there are common applications of AI-driven DMAIC that are based on: AI-driven Tower Maintenance: According to Uluskan and Karşı (2022), predictive maintenance algorithms identify the weakening of tower parts by observing the vibration, temperature, and electrical trends. Customer Churn Prediction: Bukhari and Akhtar (2024) explain how telecom companies apply machine-learning classification models to detect high-risk customers, i.e., those most likely to leave, and apply to them specific retention strategies. Comparison of the Data against another one: Maged, Haridy, Awad and Shamsuzzaman (2023) demonstrate the use of clustering and deep learning methods to identify irregular spikes in traffic with regard to fraud, outages or external disturbances. Optimization of the Quality of the service in Dense Urban Networks: Mansour, Abohashima and Elkhouly (2025) reveal that real-time AI models adapt the power, frequencies and handover thresholds in order to minimize call-drops in congested locations. These cases confirm the notion that the introduction of AI to the DMAIC can significantly improve the analytical level, speed, and accuracy of process enhancement in the context of telecom activities. 6. AI ENHANCED DMAIC ADVANTAGES IN TELECOMMUNICATIONS. Incorporating Artificial Intelligence into the DMAIC model brings significant benefits to the telecommunications company so that it can shift towards predictive, automated and highly efficient quality management. The transition of the conventional tools of analysis to the AI-based one, as described by Bukhari and Akhtar (2024), increases the speed, accuracy, and trustworthiness of the projects in Six Sigma. This section shows the most important advantages that appear in research and industry applications. 6.1 Rapid and More Nimble Root-Cause Analysis. The AI enhances the analytical strength of the Analyze phase to a considerable extent by automating the process of unearthing complicated relationships in telecom datasets. Escobar, Macias and McGovern (2022) assert that the conventional tools are not usually able to identify non-linear patterns that impact network reliability. Machine learning is able to counter this limitation by determining the hidden variables and other subtle hints of failures or performance deterioration. Mansour, Abohashima and Elkhouly (2025) also point to the fact that real-time AIs will be able to identify faults within seconds, as opposed to hours or days of analysis with manual techniques. The insight allows to make faster decisions and shortens the time of the entire DMAIC cycle. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [347] 6.2 Improved Predictive Quality Control. One of the most disruptive opportunities of AI to telecom operators is the prediction capabilities. According to Santacruz (2023), machine-learning models have the ability to predict network failures, indicate disruption and machine malfunctions before they impact the customers. In the article by Uluskan and Karşi (2022), predictive six sigma enables companies to move beyond reactive trouble-solving and prevention. With the ability to predict quality concerns in advance, telecom operators can save on downtime, prevent service failures and have a more consistent performance throughout their networks. 6.3 Automation of Real-Time Monitoring and Processes. Measure and Control phases are enhanced by AI as it allows telecom systems to be constantly and automatically monitored. The article by Maged, Haridy, Awad and Shamsuzzaman (2023) demonstrates that the ML-based dashboards will provide real-time display of such performance indicators like traffic flows, interference levels and tower health indicators. Equally, Mansour, Abohashima and Elkhouly (2025) show that AI-driven notifications initiate immediate responses to deviations in performance of a predetermined target threshold. This automation reduces the human error, shortens a response time and decreases the potential of defects escalating. 6.4 Better Customer Experience. The reliability and responsiveness of telecom services is closely related to customer satisfaction. AI-empowered DMAIC helps go directly to customer experience improvement by enhancing the quality of services and providing them with a feeling of personal care. According to Sood and Dhull (2024), NLP allows focusing on the customer complaints and expectations more precisely, and the company may adjust its improvement projects accordingly. Kabeer (2025) further notes that predictive churn models will assist telecom operators in detecting unhappy customers in good time, so that a personalized approach to retention can be formed based on the collected data. AI-enhanced DMAIC will improve the company loyalty and churn by increasing the reliability of the services and creating a proactive intervention possibility. 6.5 Increased Operating Efficiency and Reduction of the Cost. Automation and foreseeability of AI can save a lot of operational cost. As Channappagoudar, Bhat and Gijo (2025) point out, Industry 4.0 technologies simplify the process of data collection, save manual labor, and limit the number of interventions that are not needed in the network. According to Escobar, Macias and McGovern (2022), reinforcement learning and optimization algorithms allow telecom companies to allocate resources more efficiently, decreasing the level of waste and increasing the level of utilization. In the case of Uluskan and Karsi (2022), predictive maintenance is the most effective approach to reducing the cost of repair because tasks are performed on equipment before the problems develop into failures. In general, AI can assist telecom operators to ensure that the service is of high quality and the long-term costs of operation are minimized. 6.6 More Powerful Decision-Making and strategic planning. The adoption of AI in DMAIC increases the predictive ability of telecom managers since it offers more comprehensive insights into the tendencies of performance, customer behaviours, and risks in the processes. Bukhari and Akhtar (2024) state that the AI enhances data-driven decision-making by eliminating subjectivity and aiding the evidence-based analysis. Maged, Haridy, Awad and Shamsuzzaman (2023) prove that the AI-based performance projections assist organizations with planning the upgrades, and allocating the bandwidth and controlling network growth in a more efficient manner. This will result in smarter strategic planning and resource management over the long term. Volume-09 Issue 11, November-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research & Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [348] 7. CHALLENGES AND LIMITATIONS Even though AI-emerging DMAIC is highly beneficial in telecommunications companies, the journey is not devoid of challenges. Telecom operators have to overcome technical, organizational and ethical obstacles before becoming fully Quality 4.0 before it is fully matured as the case has been brought out in numerous research papers. These constraints have an impact on both the speed and effectiveness and sustainability of AI-based Six Sigma projects. 7.1 Data Quality and Integration Problems. Making sure that quality and reliable data is available is one of the most basic challenges. Maged, Haridy, Awad and Shamsuzzaman (2023) point out that telecom datasets typically include noise, missing values or inconsistencies that add complexity to the training of AI models. There might be incomplete network logs, equipment sensors can be faulty and customer information can be split across systems. According to Channappagoudar, Bhat and Gijo (2025), there is additional complexity in integrating data between more than one OSS/BSS platform, the IoT devices and older infrastructures. The absence of regular data governance can cause AI models to make incorrect predictions, which will result in untrustworthy decisions to improve. 7.2 Skills Impact and Job Preparedness. The application of AI to DMAIC necessitates the development of new capabilities that would integrate data science and Six Sigma skills. According to Bukhari and Akhtar (2024), a significant number of telecom employees have statistical quality tools experience, but they have not been trained with machine learning algorithms, model tuning, or data engineering. According to Sood and Dhull (2024), organizations usually have difficulty putting together cross-functional teams that are able to handle AI technologies and process improvement frameworks. The lack of hybrid talent will reduce the adoption of AI and raise the rate of reliance on outside consultants.