Businesses are moving beyond basic generative artificial intelligence (AI) tools towards systems that can make decisions, coordinate workflows and deliver measurable business outcomes. At TechSparks 2026, Amazon highlighted the growing role of agentic AI in enterprise operations, while CleverTap urged marketing teams to evaluate AI investments based on their impact on revenue, customer engagement and profitability.
The discussions reflect two important developments in enterprise AI: automating complex business processes and making customer marketing more responsive to individual behaviour.
Amazon Sees Agentic AI Transforming Enterprise Operations
Rajeev Rastogi, Vice President of Machine Learning at Amazon, described how agentic AI is changing the way enterprise software operates.
Unlike conventional software that follows predefined instructions, AI agents can interpret natural-language requests, develop plans and use specialised tools to complete tasks with less human intervention.
Amazon already applies these capabilities in services such as Rufus, its conversational shopping assistant. The system helps customers explore products, compare options and navigate shopping decisions. AI-powered seller assistants can also support merchants with product listings, inventory monitoring and export-related processes.
Rastogi also emphasised that generative AI is not the best solution for every business problem. Traditional machine learning models can remain more efficient for high-volume, low-latency tasks, while smaller, specialised models may reduce costs for particular applications.
For enterprises, the opportunity lies in combining different AI technologies to automate workflows without making systems unnecessarily expensive or complex.
CleverTap Urges Marketers to Measure AI by Business Results
At the same event, Anand Jain, Co-Founder and Chief Marketing Officer of CleverTap, challenged businesses to look beyond using AI simply to produce content faster.
Although AI can accelerate email drafting, creative development and document analysis, productivity gains do not automatically translate into increased revenue or improved customer retention.
Jain argued that marketing technology must progress towards systems that make decisions based on individual customer needs and real-time behaviour.
This approach changes the focus from sending more messages to determining which interaction is most likely to create value for a particular customer.
For example, rather than sending the same discount to every user, a marketing platform could consider browsing behaviour, previous purchases and current activity before selecting an appropriate message or offer.
Real-Time Decisioning Becomes a Priority
CleverTap identifies three broad stages in marketing AI maturity:
Productivity: Using AI to speed up routine marketing tasks and content creation.
Experimentation: Testing different messages, offers and customer experiences to understand what performs best.
Decisioning: Selecting and adapting the most appropriate customer interaction based on individual behaviour and business objectives.
The third stage requires more than automation. It involves combining customer history with current activity, evaluating possible actions and learning from the results.
CleverTap's approach to real-time personalisation focuses on choosing the next action for each customer while operating within predefined business and brand guidelines. <
Why Marketing Teams Need Clear AI Performance Metrics
As businesses invest in AI-powered marketing platforms, they need to assess whether these systems generate incremental value rather than simply increase output.
Useful performance measures include:
Conversion rates: Whether AI-driven experiences encourage more customers to complete a purchase or desired action.
Customer retention: Whether personalised engagement helps customers remain active over time.
Customer lifetime value: Whether marketing improves the long-term value generated by customer relationships.
Incremental revenue: Whether an intervention creates additional business results compared with what would have happened without it.
Cost efficiency: Whether improved outcomes justify technology, implementation and operating costs.
These metrics help businesses distinguish between tools that make teams faster and systems that improve commercial performance.
Human Oversight Remains Essential
Greater autonomy also introduces new responsibilities. AI agents may interact with multiple systems, access business information and initiate actions that affect customers or operations.
Enterprises must therefore establish clear limits around budgets, data access, brand communications, regulatory compliance and approval requirements. Human oversight remains important for sensitive decisions, unusual cases and activities with significant financial or reputational consequences.
A well-designed AI strategy should balance autonomy with accountability, ensuring that systems operate within defined objectives and can be evaluated against measurable outcomes.
Conclusion
Amazon's enterprise AI vision and CleverTap's marketing strategy discussion point towards a common shift: businesses are moving from AI-assisted tasks to AI systems that can execute workflows and make decisions.
For enterprise leaders, success will depend on choosing the right technology for each task, controlling operating costs and measuring actual business impact. For marketers in particular, the next phase of AI adoption will be defined not by how much content they produce, but by how effectively they improve customer experiences, retention and profitability.
