Artificial Intelligence (AI) has evolved from a niche domain of academic research to a transformative force across industries. In this insightful discussion between James Jude and Koshel Karpati, both veterans in AI innovation, we explore the nuanced applications of Generative AI (GenAI) and Agentic AI in business. This article distills their conversation, focusing on actionable insights for business leaders, as well as forward-looking predictions about how these technologies will continue to shape enterprise operations.
The Evolution of AI: Data, Compute, and Algorithms
Koshel Karpati, a seasoned AI expert, reflected on the evolution of AI over the past few decades. He highlighted three pivotal factors that have fueled this transformation:
- Data Growth: AI models thrive on vast amounts of data. Over the last 15 years, the ability to feed massive datasets into models has enabled them to recognize patterns and generate insights at unprecedented levels.
- Compute Power: The rise of high-performance computing, particularly GPUs, has accelerated AI's ability to process data and train models faster.
- Algorithmic Innovation: Breakthroughs like the Transformer architecture (introduced in 2017) have optimized how models learn, allowing for greater efficiency and adaptability even with smaller datasets.
These advancements have brought AI to the forefront of business innovation, creating opportunities for automation, decision-making, and personalization at scale.
Generative AI: Automating Tasks and Enhancing Productivity
Breaking Down Generative AI’s Capabilities
Generative AI has become synonymous with creating human-like text, images, and even videos. However, as James and Koshel noted, its true business value lies beyond these surface-level applications. Generative AI can automate repetitive tasks, enhance customer interactions, and create tailored content.
One tangible example shared was in customer service, where GenAI can generate case wrap-up summaries based on automated transcripts. This use case alone has demonstrated:
- A 50% reduction in time spent on post-call documentation.
- Increased consistency and standardization of case notes.
- More focus for agents to engage in strategic or complex tasks.
In marketing, GenAI has enabled the creation of personalized content at scale, generating tailored images and messaging for specific audiences. However, the technology’s susceptibility to hallucinations - producing incorrect or biased outputs - has tempered its use in critical decision-making scenarios.
Task Automation vs. Job Automation
Koshel emphasized that while GenAI excels at automating specific tasks, it has yet to replace entire job roles. For example, it can assist in note-taking during meetings or patient-doctor conversations, but human oversight remains critical to ensure accuracy and context.
However, the long-term trajectory of AI suggests that as models become more reliable and capable, they may begin to handle end-to-end workflows, fundamentally altering job functions.
Agentic AI: The Next Frontier
What is Agentic AI?
Agentic AI refers to systems that not only process tasks but also exhibit agency - the ability to plan, make decisions, and take actions autonomously within a defined context. Unlike GenAI, which predominantly generates outputs based on prompts, Agentic AI operates with a higher level of independence.
Key features of Agentic AI include:
- Environment Awareness: Understanding the business context and external variables.
- Feedback Integration: Adjusting actions based on continuous input from users or systems.
- Tool Utilization: Using external resources such as databases, search engines, or APIs to achieve goals.
- Planning Capability: Mapping out multi-step processes to achieve defined objectives.
For instance, an agent could plan an itinerary for a business trip, factoring in hotel availability, flight schedules, and budget constraints. Unlike GenAI, which might provide isolated recommendations, an agent connects these elements into a cohesive plan.
Business Implications of Agentic AI
While Agentic AI holds significant promise, businesses remain cautious about relinquishing decision-making to AI systems. Current applications often focus on generating recommendations or draft actions that require human approval.
Koshel predicts that as models become more adept at understanding complex workflows, organizations will begin to trust AI with low-stakes decisions. However, accountability will remain a critical consideration, especially for decisions with significant financial or legal implications.
Key Challenges for Enterprises
Data Governance and Security
For Chief Data Officers (CDOs) and Chief Analytics Officers (CAOs), ensuring data security and governance is paramount when integrating AI. Koshel highlighted several key concerns:
- Data Permissions: AI systems must respect role-based access to sensitive information.
- Zero Data Retention Policies: Vendors need to guarantee that enterprise data is not retained or used to train external models.
- Guardrails Against Hallucination: AI outputs must be monitored to prevent the dissemination of incorrect or incomplete information.
Avoiding Redundancy
A notable pitfall for enterprises is using AI to solve problems that existing tools already address. James shared an example where a company applied GenAI to analyze sales data, only to realize they were duplicating efforts already handled by their business intelligence tools. This underscores the importance of orchestration - choosing the right AI tools for the right tasks.
The Future of AI: Predictions for 2026
Koshel offered several predictions for how AI will evolve in the next 18 months:
- Composability: AI systems will become modular, allowing businesses to combine specialized models for more tailored solutions.
- Task-Specific Models: Smaller, more cost-efficient models will handle simpler tasks, while advanced reasoning models tackle complex problems.
- Education Reform: As AI automates entry-level roles, academic institutions will need to realign curricula to prepare students for emerging careers in AI-enabled fields.
- Global Power Shifts: The widespread adoption of AI will reshape global industries, driving competition and innovation across borders.
- Ambient AI: With AI becoming cheaper and smaller, we can expect its integration into everyday devices, from smartphones to household appliances.
Key Takeaways
- Generative AI excels in task automation, such as summarization and content creation, reducing time spent on repetitive work.
- Agentic AI brings decision-making capabilities, planning multi-step processes and adapting to feedback, but trust and accountability remain barriers to adoption.
- Data governance is critical when deploying AI, including enforcing access controls, zero data retention policies, and safeguards against hallucinations.
- Education systems must adapt to prepare the workforce for an AI-dominated future, focusing on skills like prompt engineering and data literacy.
- AI adoption requires orchestration, ensuring the right tools are applied to the right problems without redundancy or overspending.
Conclusion
The rapid evolution of GenAI and Agentic AI signals a transformative shift for businesses. While the technology has already demonstrated its ability to enhance productivity and automate tasks, its full potential lies in reshaping entire workflows and decision-making processes. For executives and decision-makers, the challenge is twofold: leveraging AI responsibly and preparing their organizations for the profound changes it will bring.
By understanding these technologies and their applications, business leaders can position themselves at the forefront of innovation, driving efficiency, scalability, and competitive advantage in a rapidly evolving market landscape.
Source: "Unleashing Agentic AI: Transforming business processes with Generative AI" - Wipro, YouTube, Aug 14, 2025 - https://www.youtube.com/watch?v=eGbJu-qn8Rg
Use: Embedded for reference. Brief quotes used for commentary/review.
.png)