By Kuljesh Puri, Senior Vice President and General Manager – Communications, Media & Technology, Persistent Systems
Smart devices have evolved beyond mere connectivity. Today, they are intelligent, adaptive
systems that anticipate user needs, learn from behaviour, and respond in real time with
human-like precision. Fuelled by artificial intelligence (AI), these devices are transforming
consumer, industrial, and enterprise sectors – enhancing user experience, reducing latency,
strengthening security, and optimizing network efficiency.
Data processing lies at the core of these intelligent devices, enabling them to incorporate
complex AI models like deep and recurrent neural networks, along with other advanced
machine learning (ML) architectures. These systems continuously analyze vast data
streams, adapting to user interactions and environmental cues. For example, a smart
telecom system can optimize bandwidth allocation based on real-time demand, while IoT-
enabled industrial sensors can pre-empt equipment issues. Innovations in model
optimization and parameter tuning are accelerating AI’s growth. Global Market Insights
estimated the global AI in telecommunication market was worth USD 2.7 billion in 2024 and is projected to achieve a compound annual growth rate (CAGR) of 32.6 percent from 2025 to 2034.
Real-time Processing with Edge AI
For AI-enabled applications to operate in a hyperconnected ecosystem, they require low-
latency computing power, something that conventional cloud architectures can’t offer. They
are often hindered by network congestion and processing delays, making them less viable
for real-time analytics and mission-critical operations – such as network traffic management, predictive maintenance, and dynamic bandwidth optimization within 5G networks. Edge AI overcomes these limitations by decentralizing computational workloads and processing data at the source rather than relying on distant cloud servers.
The architectural shift allows immediate response times, reduces data transmission
overhead, optimizes bandwidth utilization, and enhances security by restricting access to
external networks. As per the IMARC group, India’s edge computing market reached USD
567.3 million in 2024 and is projected to grow at a CAGR of 15.7 percent from 2025 to 2033, reaching approximately USD 2.37 billion by 2033. These projections highlight the significant role of edge computing in supporting AI-native infrastructure for telecom and intelligent IoT ecosystems.
Reshaping The Telecommunications Sector with AI-driven Personalization
Another reason IoT devices are gaining prominence is their ability to personalize customer
experiences. As AI-driven systems proactively learn user behaviour to anticipate needs,
optimize interactions, and enhance convenience, they are rewriting the way users interact
with technology. Whether it’s a home automation system that adjusts lighting and climate
based on personal preferences or virtual assistants that refine responses over time, these
technologies are creating more intuitive, responsive, and seamless user experiences. In the
telecommunications sector, AI-driven systems are used to analyze customer usage patterns, predict their needs, and offer personalized service plans.
A notable example is how a major U.S. telecom provider enhanced its GCP-hosted
conversational AI model using a GCP-native data platform to process over 1 billion
messages daily. This enabled real-time personalized assistance, reduced churn, and
boosted customer satisfaction through actionable insights prompted based on caller
sentiment and service performance. Such advancements illustrate how AI is redefining
telecom customer relationship management (CRM) and driving engagement.
Securing Data and Optimizing Infrastructure
With the rise of 5G-enabled IoT devices, there is a staggering volume of data generated,
which brings with it significant challenges related to privacy, potential data breaches, and the urgent need for real-time threat mitigation. Organizations can leverage edge computing or build a multi-layered security framework that strikes a balance between computational
efficiency and innovative AI solutions to address these cyber risks.
The security frameworks can also be strengthened through technologies such as end-to-end encryption, zero-trust architectures, and AI-integrated anomaly detection. One can even enable continuous AI learning by integrating federated learning and differential privacy techniques to safeguard sensitive data.
The smooth functioning of the IoT ecosystem hinges upon network performance.
Incorporating tools like AI-powered load balancing, dynamic resource allocation, and
network slicing can maximize efficiency and performance. Complementing techniques such
as model compression, adaptive AI inference, and distributed processing can contribute to
seamless data flow, minimal latency, and efficient energy usage across smart networks.
As emerging technologies continue to shape the next generation of smart devices, real-time connectivity and intelligent automation are transforming how telecoms engage with users.
This evolution transcends technology, becoming a catalyst that unites people, ideas, and
industries. By seamlessly integrating these innovations, we are enhancing device
capabilities and creating a more adaptive IoT ecosystem.
As these devices grow more context-aware and predictive, the lines between digital and
physical experiences will blur, unlocking new opportunities for personalization, efficiency,
and inclusion. Ultimately, the future belongs to those who can harness the synergy of data,
AI, and connectivity to create a meaningful, responsible, interconnected, and intelligent
future.
