Akamai Summit: AI Implementation Challenges in India
Tech leaders discuss AI implementation challenges at an Akamai summit in India.
Key Highlights
Akamai’s Digital Leadership Summit in Bengaluru addressed AI implementation challenges: inference costs, voice AI, API security, and sovereign models in India.
The Akamai Digital Leadership Summit recently convened in Bengaluru to explore the practical realities of deploying artificial intelligence at scale in India. The summit brought together 125 senior technology leaders to discuss key challenges such as managing inference costs, implementing voice AI, ensuring API security, and developing sovereign models.
Sumant Narayanan, Akamai’s Regional Director for India and SAARC, emphasized the shift from focusing on the size of foundation models to how enterprises can use them to deliver customer value. He also noted the increased security risks associated with AI tools.
Akamai CTO Dr. Robert Blumofe traced the company’s AI journey from the deep learning breakthrough of 2012 to the present day. He cautioned against becoming overly reliant on large language models (LLMs) and stressed the importance of using diverse AI and machine learning techniques.
Jigar Halani from NVIDIA highlighted India’s unique challenges, including the need to provide services at scale with minimal transaction costs. He cited examples such as the Bhashini platform, which handles millions of daily API calls in multiple Indian languages, and AI systems that improve biometric authentication and fraud detection.
A panel moderated by Hrishikesh Varma of Akamai discussed the specifics of running AI at scale across different organizations. Sanath Moguluri of Reliance Jio described the latency challenges of serving millions of users across various devices. Sagar Gaonkar, CTO of Eloelo, explained their approach to live content moderation, which separates edge and cloud decisions. Pranav Tiwari from Postman discussed how AI is transforming the connectivity layer, with AI agents increasingly involved in application connectivity.
A fireside chat between Mohit Saxena, Co-Founder and CTO of InMobi and GlanceAI, and Adam Karon, Akamai’s COO, addressed the costs of AI implementation in India. Saxena noted the need to reduce costs significantly to make AI services viable for the Indian market, emphasizing the importance of efficient query handling and specialized models.
Kiran Kumar Katreddi, VP of Platform Engineering at Meesho, discussed the engineering constraints of serving over 200 million users in Tier-III and Tier-IV towns. He highlighted Meesho’s development of its own Bharat ML stack to handle the demands of its zero-commission business model.
Vijay Kolli, Akamai’s Regional VP for Enterprise Security, focused on the security challenges posed by the expanding attack surface of AI systems. Mukesh Solanki, CISO at KreditBee, and Sujatha Iyer, Head of AI Security at Zoho Corp, discussed the increasing sophistication of hackers using AI tools and the importance of embedding security into software development from the outset.
Ganesh Gopalan, CEO of Gnani.ai, advocated for sovereign AI, where enterprises retain ownership of their AI systems’ intelligence. He emphasized the importance of owning every layer of the voice stack to protect customer data, improve user experiences, and control costs.
The summit underscored that the next phase of AI adoption will depend on engineering discipline and the ability to run AI systems reliably and cost-effectively. The discussions highlighted the shared responsibility in building and deploying AI systems at scale.