Client Introduction
Dell Technologies is a global leader in information technology, providing a broad portfolio of infrastructure, software, and services to enterprises, public sector organisations, and consumers worldwide. Headquartered in Round Rock, Texas, the company operates across more than 180 countries and serves millions of customers at scale.
Dell’s business model spans direct and channel sales, delivering products and services that include servers, storage, networking, end-user computing devices, and a comprehensive suite of IT support and managed services. Within its services division, Dell manages a vast and complex support ecosystem responsible for handling millions of warranty claims and technical support interactions annually.
Dell Technologies consistently ranks among the world’s largest technology companies by revenue. In 2024, the company was recognised as the world’s number one provider of AI infrastructure. [1]
Project Introduction
Project Name: Telemetry, Data Science and AI Enablement for Warranty Cost-Out Initiatives
Department: Services AI Product Management and Strategy
Role: Product Owner
The core objective of this initiative is to reduce the cost of delivering technical support under warranty, while simultaneously improving the quality and speed of support outcomes for customers and internal teams.
Dell’s support operations generate enormous volumes of device telemetry, case data, and support interaction signals daily. Without an intelligent framework to orchestrate and act on this data in real time, support teams face inefficiencies including manual triage, delayed resolution, and suboptimal routing of cases. These inefficiencies translate directly into elevated support costs and longer resolution times.
This project addresses those challenges by delivering a platform centred on real-time telemetry orchestration and AI-driven insights. The platform produces Next Best Action (NBA) recommendations that guide support engineers and automated systems toward the most effective resolution path for each support case.
The initiative also establishes the foundational architecture required for future Agentic AI systems, enabling autonomous decision-making and workflow automation within Dell’s support operations.
Scope of work includes:
- Platform design and roadmap ownership
- Real-time telemetry data pipeline development and orchestration
- AI and machine learning model integration for triage and remediation recommendations
- Next Best Action engine design and deployment
- Signal quality improvement and data governance
- Stakeholder alignment across engineering, product, data science, and operations teams
- Laying the technical foundations for autonomous, agentic AI workflows
Technical Implementation
Architecture and Technology Stack
The platform is designed around a real-time data ingestion and orchestration architecture. Device telemetry from Dell hardware is collected, processed, and enriched with contextual support data before being passed to AI inference layers.
Dell’s approach to telemetry has been central to its AI strategy across multiple product lines. Customers connected to Dell’s SupportAssist technology can leverage telemetry and AI to address hardware issues without human intervention, including autonomously correcting errors across entire device fleets. [2] This same telemetry-first philosophy underpins the warranty cost-out platform described in this case study.
Key components include:
- Real-time telemetry ingestion pipelines capable of processing high-velocity device signals
- A centralised data platform that aggregates telemetry, case metadata, and historical resolution data
- Machine learning models trained on support interaction data to power triage, classification, and remediation recommendations
- A Next Best Action engine that surfaces prioritised, context-aware recommendations to support engineers and automated workflows
- Data quality and signal validation layers to ensure accuracy and reliability of inputs to AI models
To effectively train AI models at this scale, Dell worked across teams to standardise telemetry data formats, enabling all product information to feed into a unified intelligence layer. [3]
Solution Design
The solution is designed to intervene at the point of case triage, before manual effort is expended on investigation paths that AI can either automate or accelerate.
Over an 18-month development period, Dell’s team designed the NBA engine to replicate and ultimately enhance human agent behaviour. The system integrates multiple AI capabilities to aggregate and analyse content from technical sources, provide step-by-step guidance for faster issue resolution, tailor recommendations to each product line, and continuously learn from agent feedback in real time. [3]
Key design principles include:
- Smarter triage: AI models assess incoming support cases against telemetry signals to classify severity, likely root cause, and recommended resolution path before a human engineer engages
- Automated remediation paths: Where confidence thresholds are met, the system triggers automated fixes or guided resolution workflows, reducing the need for manual intervention
- Improved signal quality: Data pipelines are designed to clean, validate, and enrich telemetry data, ensuring AI models operate on high-quality inputs
- Human-in-the-loop design: Where AI confidence is lower, the system surfaces ranked recommendations to engineers rather than acting autonomously, preserving human oversight
- Agentic AI foundations: The architecture introduces autonomous workflow triggers and data-driven decision frameworks, positioning the platform as the base layer for future fully autonomous support operations
Development and Execution
The product is delivered using an agile product management methodology, with the Product Owner responsible for roadmap prioritisation, backlog management, and stakeholder alignment.
AI success in this initiative is not confined to one department. Dell emphasised cross-functional collaboration to reduce duplication and align teams around shared objectives. Keeping leadership informed about AI initiatives ensures alignment, funding, and continued progress. [3]
Key execution phases include:
- Discovery and data assessment: Evaluation of available telemetry signals, case data quality, and existing support workflows to identify high-impact AI intervention points
- Platform foundation build: Establishment of core data pipelines, telemetry ingestion infrastructure, and model serving architecture
- NBA engine development: Design, training, and validation of AI models that produce actionable Next Best Action outputs
- Integration and deployment: Embedding NBA recommendations into existing support tooling and workflows used by support engineers
- Signal quality improvement: Ongoing iteration on data validation, enrichment, and feedback loops to improve model accuracy over time
- Agentic AI roadmap: Definition of future capability milestones toward autonomous workflow triggers and decision automation
Cross-functional stakeholder groups involved include Services AI Product Management, Data Science, Engineering, Support Operations, and Finance.
Performance and Outcomes
By connecting product telemetry with service operations, Dell ensured that AI insights directly informed design, support, and delivery. Dell initially focused on using AI to improve customer experience and reduce costs, rather than treating AI as a product to be sold in isolation — a strategy intended to build long-term value and operational efficiency. [3]
The platform is designed to deliver measurable improvements across the following dimensions:
- Reduction in manual intervention: By enabling automated remediation and AI-guided triage, the platform reduces the proportion of cases requiring full manual engineering effort
- Improved decision-making speed: Support engineers receive context-rich, ranked recommendations at the point of case intake, reducing time spent on investigation
- Higher signal quality: Cleaner, validated telemetry data improves the accuracy and reliability of AI model outputs over time
- Cost-out realisation: Each reduction in manual handling and improvement in first-contact resolution translates directly to warranty cost savings at scale
- Foundation for future automation: The platform establishes the data infrastructure and decision logic required for Agentic AI, enabling future autonomous case resolution
Note: No verified public quantitative KPI data is available for this initiative at this time. Specific performance metrics should be provided by the project team for inclusion prior to publication.
Business Impact
Operational efficiency: Reducing manual triage and intervention steps lowers the cost per support case and frees engineering capacity for higher-complexity work.
Cost-out at scale: Given the volume of warranty cases Dell manages globally, even incremental improvements in automation rates and resolution accuracy represent material cost reductions. Dell’s broader enterprise AI research indicates that on-premise AI inference deployments can deliver significantly lower total cost of ownership at scale compared to cloud-only alternatives, reflecting the company’s commercial focus on cost-efficient AI delivery. [4]
Strategic positioning: The platform positions Dell’s Services division to operate with AI-native workflows, aligning with the company’s broader commitment to AI-driven innovation. Dell’s Agentic AI deployments in adjacent business units, such as its telecom division, have demonstrated the ability to reduce manual work and optimise costs through autonomous workflow orchestration. [5]
Future readiness: By establishing the architectural foundations for Agentic AI, this initiative ensures Dell is positioned to deploy autonomous support capabilities as AI technology and enterprise risk appetite matures. The first release of Dell’s Data Analytics Engine Agentic Layer is scheduled for early 2026. [6]
Customer experience: Faster, more accurate support resolutions benefit end customers directly, supporting Dell’s service quality commitments and customer satisfaction targets.
Appendix: Glossary of Terms
Agentic AI: A category of artificial intelligence system that can autonomously plan, make decisions, and execute multi-step actions without requiring a human to direct each step. In this context, it refers to future AI systems capable of resolving support cases end-to-end without human intervention.
AI (Artificial Intelligence): The use of computer systems to perform tasks that typically require human intelligence, such as recognising patterns, making decisions, and generating recommendations.
API (Application Programming Interface): A defined interface that allows two software systems to communicate and exchange data with each other.
Backlog: In agile product management, a prioritised list of work items, features, or tasks that the development team will address over time.
Cloud Infrastructure: Computing resources — including servers, storage, and networking — delivered over the internet by a cloud service provider, enabling scalable and flexible technology operations.
Data Pipeline: A series of automated processes that move, transform, and enrich data from one system to another, ensuring it is ready for analysis or use by AI models.
Data Science: The discipline of extracting insights and knowledge from structured and unstructured data using statistical analysis, machine learning, and data engineering techniques.
First-Contact Resolution: A support performance metric measuring the percentage of cases resolved during the customer’s first interaction, without requiring follow-up contact.
KPI (Key Performance Indicator): A measurable value used to assess how effectively an organisation or system is achieving a defined objective.
Machine Learning (ML): A subset of artificial intelligence in which computer systems learn from data to improve their performance on specific tasks over time, without being explicitly reprogrammed.
NBA (Next Best Action): An AI-driven recommendation framework that identifies and presents the most appropriate next step for a support engineer or automated system to take, based on available data and context.
Product Owner: In agile development, the individual responsible for defining the vision, prioritising the product roadmap, managing the backlog, and ensuring the team delivers maximum value.
Remediation: The process of identifying and resolving the root cause of a technical problem, restoring a system or device to normal operation.
ROI (Return on Investment): A measure of the financial return generated relative to the cost of an investment, typically expressed as a percentage.
Signal Quality: The accuracy, completeness, and reliability of data inputs used by AI models. Higher signal quality leads to more accurate and trustworthy model outputs.
Stakeholders: Individuals or groups with an interest in, or influence over, the outcome of a project. In this context, this includes product, engineering, data science, operations, and finance teams.
Telemetry: Automated collection and transmission of data from remote devices or systems. In this context, telemetry refers to diagnostic and operational data transmitted by Dell hardware to support monitoring and analysis.
Triage: In technical support, the process of assessing and prioritising incoming cases based on severity, complexity, and urgency, to determine the appropriate response path.
Warranty Cost-Out: A business objective focused on reducing the total cost of fulfilling warranty obligations, including the cost of support labour, parts, and case handling.
References
- Dell Technologies Newsroom. Dell AI Data Platform Advancements Unlock the Power of Enterprise Data to Accelerate AI Outcomes. October 2025. View source
- Dell Technologies Newsroom. Dell Technologies Helps Organizations Create a Modern Workplace with New AI Experiences. February 2024. View source
- TSIA. Case Study: Inside Dell’s AI Strategy for Predictive and Proactive Support. November 2025. View source
- Dell Technologies Newsroom. Dell Technologies Fuels Enterprise AI Innovation with Infrastructure, Solutions and Services. May 2025. View source
- Dell Technologies Blog. Dell Unveils Scalable AI Solutions for Telecom. October 2025. View source
- Dell Technologies Newsroom. Dell AI Data Platform Advancements Unlock the Power of Enterprise Data to Accelerate AI Outcomes. October 2025. View source
Ready to Explore AI-Driven Support Solutions?
Are you exploring AI-driven solutions for your support operations? Get in touch to learn how intelligent telemetry and Next Best Action frameworks can transform your service delivery.
Pramod Siddabattuni - Product Consultant / Functional Solution Architect 