How ChatGPT and Generative AI Work Under the Hood – From Tokens to Transformers

Event Date: 14th June 2025 

Event brief description 

This The expert talk titled "How ChatGPT and Generative AI Work Under the Hood – From Tokens to Transformers" provided an in-depth exploration of the inner mechanisms powering modern AI language models. The session walked participants through the complete journey of how text is processed—from tokenization to the advanced transformer architecture that underlies models like ChatGPT. 

The speaker effectively demystified complex AI concepts by explaining the roles of neural networks, attention mechanisms, and large-scale data training. Attendees gained valuable insights into how ChatGPT learns, generates responses, and adapts to varied inputs, along with real-world examples showcasing its applications across industries. 

The session also highlighted the technical and ethical challenges in deploying such models, fostering a deeper understanding of responsible AI use. Students, educators, and professionals found the talk highly enriching, as it bridged theoretical knowledge with practical implementations. 

Overall, the expert talk successfully enhanced the audience’s understanding of Generative AI, equipping them with foundational knowledge of the technology that powers many of today’s intelligent applications. 

Event Detailed Description

An expert talk titled "How ChatGPT and Generative AI Work Under the Hood – From Tokens to Transformers" was successfully conducted on 14th June 2025 at Galgotias University. The session was led by Mr. Akash Shah, an AI and Generative AI Consultant at Tata Consultancy Services (TCS), who brought with him valuable industry insights and practical expertise on the fast-evolving domain of artificial intelligence. 

The talk was attended by enthusiastic BCA and MCA students, eager to explore the workings of Generative AI models and understand their real-world implications. The primary objective of the session was to expose students to cutting-edge AI concepts, bridge the gap between academic knowledge and industry trends, and inspire them towards innovation and career growth in the AI domain.

Mr. Shah began the session by breaking down complex concepts behind models like ChatGPT, emphasizing how language is processed, tokenized, and generated by transformer based architectures. He elaborated on the role of training data, attention mechanisms, and model tuning in generating human-like responses. 

One of the highlights of the session was the discussion on real-world use cases of Generative AI, including applications in customer service, healthcare, content creation, and education. Mr. Shah also encouraged students to explore project ideas that leverage these technologies, giving them a roadmap to apply their classroom learning in practical domains. 

The session also addressed crucial themes like Ethical AI, data privacy, and responsible AI deployment—an essential aspect of today’s AI-driven world. Students appreciated this part of the discussion, as it highlighted not just the technical side of AI, but also its social and ethical responsibilities. 

An engaging Q&A session followed, during which students asked thoughtful questions and even showcased their own AI-related projects. Mr. Shah provided detailed feedback and encouraged their curiosity and experimentation. To recognize participation and enthusiasm, goodies were distributed to students, further boosting the interactive and positive atmosphere of the event. 

The session received highly positive feedback. Students expressed that many of their doubts were resolved, and the talk helped them connect theoretical concepts with practical applications. They left the room motivated, informed, and better equipped to pursue AI related academic and professional paths. 

Overall, the expert talk proved to be an enriching learning experience, aligning perfectly with the curriculum and providing students with valuable exposure to the world of AI and its future possibilities. 

Department Name – School of Computer Application and Technology

Event Outcome

The expert talk on “How ChatGPT and Generative AI Work Under the Hood – From Tokens to Transformers” proved to be a highly impactful and enriching session for the students of BCA and MCA at Galgotias University. Conducted by Mr. Akash Shah, AI and Gen AI Consultant at TCS, the session successfully met its core objective of bridging theoretical learning with real 

world AI applications. 

Key Outcomes: 

Enhanced Conceptual Understanding: Students gained a deep understanding of how Generative AI models like ChatGPT function, including tokenization, transformer architectures, and training processes. 

Project Inspiration: The session sparked several innovative project ideas among students, motivating them to explore AI-based solutions. 

Industry Exposure: By interacting with an industry expert, students were exposed to real world use cases and the current trends shaping the AI domain. 

Clarification of Doubts: The interactive Q&A session helped students clear their conceptual doubts related to AI, machine learning, and ethical AI use. 

Student Engagement: Goodies were distributed as appreciation for active participation, creating a lively and encouraging learning environment. 

Career Awareness: Students received valuable insights into AI career paths and skillsets required to excel in this field. 

Overall, the session strengthened the university’s mission to offer industry-aligned education and inspired students to pursue practical, responsible, and impactful AI innovations in their academic and professional journeys. 

Related Goal