Beyond generative AI
The potential of AI to act autonomously, without human supervision, has always been a transformative promise. However, this kind of “Agentic AI” has been limited in enterprise use cases. According to Sam Witteveen, CEO of Red Dragon AI, two trends will change this perception in the next year and a half:
1. Agents in everything: AI agent-embedded alternatives to familiar software tools and services will become available, allowing users to interact with them in natural language.
2. Building blocks for agents: New tools and frameworks for building custom AI agents will emerge, enabling businesses to develop AI-driven strategies for different aspects of their operations.
Agentic AI represents the next phase of AI adoption for enterprises. This series of articles will explore the impact of Agentic AI on various aspects of organizations, including cybersecurity, IT administration, business operations, sales, marketing, and more. Additionally, the series will examine the evolving ethical and regulatory landscape.
Generative AI technologies, such as ChatGPT, have already been integrated into various industries, resulting in significant ROI. According to a Google Cloud study, 70% of companies have seen ROI on at least one use case. McKinsey predicts that generative AI technologies will add between $2.6 trillion to $4.4 trillion of value across business sectors and reduce the total amount of work required by employees by 50%-70%.
However, Agentic AI goes beyond generative AI by enabling systems to autonomously monitor events, make decisions, and take real actions. From embedded agents managing cybersecurity threats to marketing AIs generating hyper-personalized campaigns, Agentic AI is a paradigm shift that will have profound effects on enterprises and society.
Defining Agentic AI: generative AI fused with classical automation
Agentic AI combines classical automation with the power of modern large language models (LLMs), using LLMs to simulate human decision-making, analysis, and creative content. Unlike traditional automation, Agentic AI allows anyone who can use language to interact with the system and replaces static scripts with LLM-generated code-on-demand.
Agentic AI systems possess several properties:
1. Generation: Agentic AI systems harness the analytic and creative capacity of LLMs, using generated outputs as intermediate steps within a complex workflow.
2. Tool Calling: AI in agentic systems can call upon specific tools or APIs, querying data and triggering events according to the reasoning generated by the LLM.
3. Discovery: Agentic systems can access real-world data from various tools and data streams, allowing them to harness LLM generation to ask for data they need.
4. Execution: Agentic systems can take real-world actions, such as interacting with external systems or triggering processes, without human intervention.
5. Autonomy (Self-prompting): Agentic systems are “always on” and can loop through cycles of acting, evaluating, and planning without specific triggers.
6. Planning: Agentic systems can generate, prioritize, and manage sets of subordinate tasks to pursue an overall goal.
7. Composition: Agentic systems can assemble multiple components, such as queries or subroutines, into a cohesive action or response.
8. Memory: Agentic systems can build and maintain their own internal knowledge representations, enabling them to index, store, and retrieve information for further tasks.
9. Reflection: Agentic systems can evaluate the solutions they generate and try again if necessary, ensuring high-quality results.
Agentic AI has the potential to revolutionize various industries, including sales, marketing, cybersecurity, and IT operations.
Transforming enterprise
Agentic AI has enormous implications for organizations in every sector. While AI agents are still under development, they offer the potential for greater efficiency, reduced risk, and improved decision-making. However, challenges remain, such as the reliance on LLMs, which are prone to hallucination.
Several popular agent frameworks, including Langraph, Autogen, and CrewAI, have emerged. Throughout the series, we will explore use cases in various industries, review leading product offerings, and consider projects built with these tools and frameworks.
Agentic AI is already transforming sales by automating lead management and personalizing customer interactions at scale. Tools like Conversica and Relevance AI are offering AI-powered assistants that engage with potential leads, qualify them, and nurture prospects through the sales funnel. In marketing, platforms like Netcore’s Co-Marketer AI and Salesforce’s Agentforce deliver hyper-personalized content based on real-time data. Agentic AI is also revolutionizing cybersecurity by autonomously detecting and responding to threats. Additionally, platforms like Qovery are reshaping IT operations by automating infrastructure management.
The implementation of agentic AI requires thoughtful design, as these systems are not one-size-fits-all. Specialized AI agents and AI-enabled tools must be chosen based on specific requirements. Throughout the series, we will explore how enterprises can build these systems, the tools and platforms they can use, and the industries that will benefit most from agentic AI.
What’s Next?
Agentic AI has the potential to empower businesses to operate with greater efficiency, agility, and speed. As more robust offerings become available, the business case for adopting agentic AI will grow. However, thoughtful design and understanding of the technology’s hype and reality are crucial.
The upcoming articles in this series will delve into how agentic AI is reshaping marketing, sales, cybersecurity, customer service, and business operations. The series will also explore the emerging regulatory landscape and the importance of AI governance. Stay tuned for the future of AI-driven business.
