IBM Research, India
Title: Navigating the Landscape of AI Agents: Characteristics, Innovations, and Future Directions
Abstract The field of AI agents has gained substantial attention in recent years due to the surge in AI capabilities and the demand for automation in complex tasks. AI agents, with their ability to reason, plan, act, and interact, represent a significant advancement over traditional automation techniques. In this talk, we will explore the evolving trends and defining features of AI agents, providing a foundation for understanding their transformative potential. We will delve into cutting-edge paradigms such as ReACt, Reflection, and Reflexion agents, showcasing their unique approaches to problem-solving. Building on this, we will highlight recent advances in agentic software engineering research with demonstrations of our work on agentic NL2SQL and automated testing agents. Finally, we will discuss critical challenges regarding the reliability and robustness of AI agents, laying the groundwork for future research in this exciting field.
IIIT-Hyderabad, India
Title: Playing with Abstractions: At the crossroads of Software Architecture and Generative AI
Abstract Text is a powerful abstraction of reality, like architecture abstracts complex software systems. The advent of Generative AI specifically Large Language Models (LLMs) has set new benchmarks in understanding and generating human-like text, revolutionizing multiple sectors. In this talk, we explore some of these capabilities of LLMs to understand whether LLMs can be architect's new best friend. We begin with an overview of Large Language Models (LLMs) and exploring their role in generating design decisions through our research on leveraging GenAI for architecture knowledge management. We then explore their impact on automating component generation based on our ongoing research in the serverless context. Further, we provide insights on LLMs’ capabilities for supporting runtime architectural adaptation, highlighting our efforts on building Generative AI-powered autonomous CloudOps Copilot in collaboration with our industrial partner. The talk concludes by listing different opportunities and challenges drawing insights from our ongoing efforts involving Small Language Models as well as in architecting multi-agent framework for cloudOps domain.
Applied AI Consultant; Previously: Postman/Microsoft
Title: From Code to Cognition: How Agentic AI Unites Software 1.0 and Software 2.0
Abstract The release of GPT-3 in 2020 marked the advent of Generative AI and ushered in the "Software 2.0" paradigm, where tasks traditionally crafted through deterministic programming in formal languages (Software 1.0) were transformed into large data-driven predictive models. Initially, there was a rush to rapidly replace Software 1.0 workflows with Software 2.0 systems, as Generative AI advancements seemed poised to dominate the entire Software 1.0 stack, fueling aspirations of achieving the so-called “Artificial General Intelligence”. However, as these Software 2.0 systems scale, significant challenges emerge, including issues of reliability, unchecked autonomy, lagging contextual awareness, and the general inability to interact predictably and structurally with the real world. This has given rise to Agentic AI, marking a pivotal shift where Software 2.0 systems are increasingly incorporating foundational components from Software 1.0. By leveraging APIs for structured interaction, state machines for predictable autonomy, context-free grammars for guided outputs, and code interpretation for precise execution, Agentic AI blends the strengths of both paradigms, unlocking new frontiers in AI Research and Software Engineering. This talk explores this trend reversal through real-world case studies, major research contributions, and examples from leading Agentic AI frameworks. We’ll discuss how frontier tech labs and companies like Postman, GitHub, and Perplexity are scaling generative AI applications to millions of users by blending Software 1.0 and Software 2.0. Additionally, we’ll examine how emerging Agentic AI ventures in the “Service-as-a-Software” economy rely on Software 1.0 tooling as a core part of their architecture. By uniting code and cognition, this evolving symbiosis is paving the way for impactful, scalable autonomous AI systems.
confirmations awaited
confirmations awaited
confirmations awaited