Tech Mahindra & Intel: Crafting the next wave of AI-driven telco transformation

Intel’s optimized AI libraries, such as NumPy—used during Digital Twin training in this use case—further enhance computational speed and efficiency.

What specific business challenges were your clients facing?

Ticketing inefficiencies and capital planning delays have long hampered telcos. These issues not only degrade end-user experience and operations but also drive up costs and response inconsistencies. GenAI is new on the ground, but it’s rapidly gaining traction due to its potential to streamline such high-impact areas.

Our clients have been grappling with underutilized network capacity, inflated Capex, sub-optimal cost efficiency and slow decision-making. Our AI-powered digital twin solution provides real-time, scenario-based insights, allowing telcos to shift from reactive to proactive planning. Moreover, identifying use cases based on a client’s maturity in Autonomous Network (AN) operations is itself a complex exercise. That’s where our structured approach comes in.

Walk us through your approach to addressing these challenges.

We begin with a deep engagement to identify our customer’s pain points and align solutions with their strategic objectives—be it cost efficiency, agility, or faster time-to-market. At the core of our AI-led solution is a set of intelligent reflection agents that autonomously classify tickets, perform root cause analysis, and recommend resolutions with minimal human input. Integrated with enterprise ITSM platforms, these agents enable seamless support across email, chat, and ticketing systems, significantly improving service responsiveness.

A key enabler of this solution is its ability to perform high-speed AI inference directly on 5th Gen Intel® Xeon® processors—eliminating the need for additional GPUs. Powered by Intel® Advanced Matrix Extensions (AMX) and optimized using Intel® Extension for PyTorch (IPEX), this architecture delivers real-time ticket processing with exceptional performance-per-watt efficiency.

The 5th Gen Intel® Xeon® Processors serve as the AI backbone of our SLM-based ticketing solution and broader AI workloads. With built-in acceleration for large language model (LLM) workloads, they minimize inference latency and maximize throughput—enabling faster classification and resolution in high-volume IT environments. Building on this foundation, we provide clients with a clearly defined roadmap of automation use cases to elevate Autonomous Network maturity. Our telecom-trained LLMs and geospatial digital twin, powered by Intel Xeon® based architecture—leverage real-time network insights and business intelligence to enable rapid prototyping, cost-effective deployment, and measurable business outcomes.

What were the tangible outcomes of these initiatives?

The pilot results of our Smart Capex solution have been promising. We anticipate a reduction in project planning costs to the extent of 20% and nearly a 30% acceleration in decision-making timelines. While these figures continue to be refined as the solutions are commercialized, the early outcomes point to a significant leap in operational efficiency for telecom operators. By enabling real-time insights and predictive analytics, the solution empowers clients to allocate resources more effectively, streamline rollouts, and de-risk large-scale projects—marking a shift from reactive planning to data-driven strategic execution.

Supporting this performance is the underlying compute infrastructure, where 5th Gen Intel® Xeon® processors play a key role. Their balanced performance and power efficiency help us achieve low-latency inferencing without the need for additional hardware. In many cases, they can also be repurposed for general workloads when not engaged in AI tasks, offering greater flexibility and value in enterprise environments.

While cloud-based compute access continues to grow, on-premise deployments remain vital—especially for organizations handling critical or sensitive data. Many of our clients continue to prioritize data security, and our solution aligns well with these operational preferences while being scalable and future-ready.

How do you envisage the future of your partnership with Intel?

Our partnership with Intel is evolving into a strategic force driving the future of AI-led telecom transformation. Together, we’re building advanced solutions across service lifecycle management, predictive network intelligence, and smart capex planning—empowering telecom operators to reduce costs, improve uptime, and accelerate service delivery. Intel is deeply focused on innovation towards network transformation and AI that aligns seamlessly with our vision of creating cloud-native, programmable, and AI-first networks.

Going forward, our collaboration will expand into areas like edge computing, virtualized infrastructure, and intent-based network management. With the launch of Intel® Xeon® 6 processors—offering 2.4x more RAM capacity and up to 14x better media performance per watt—we’re well positioned to deliver high-performance, energy-efficient solutions at scale. Together, we aim to build a smarter, more connected, and sustainable telecom ecosystem that delivers faster deployments, improved network performance, and stronger ROI for our customers.

Disclaimer

Performance varies by use, configuration, and other factors. Learn more on the Performance Index site. Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. No product or component can be absolutely secure. Results have been estimated or simulated.

The pages slugged ‘Brand Connect’ are equivalent to advertisements and are not written and produced by Forbes India journalists.