Empathy Meets Intelligence: Designing Human-Centric Commerce in the Age of Agentic AI

The Rise (and Risk) of Agentic AI in Commerce

Agentic AI refers to AI systems that operate with a degree of autonomy. While traditional automation follows predefined workflows, agentic AI can plan, reason, and adapt dynamically based on changing goals and inputs.

Leading analyst firms consistently emphasize the paradigm shift towards agentic AI. Gartner predicts that by 2028, 33% of enterprise software applications will include agentic AI capabilities, while Forrester highlights the emergence of "customer-centric AI" that prioritizes human needs over technological efficiency.

In ecommerce, this includes:

• Conversational agents that answer customer questions, resolve issues, or guide shopping journeys
• AI shopping assistants that search across retailers, filter products based on complex criteria, and even complete purchases on behalf of the user
• Autonomous merchandisers that adjust search results, promotions, or page layouts in real time based on signals like demand, margin, and shopper cohort behavior

At their best, these agents remove friction, anticipate needs, and respond faster than any human could. But they also introduce new risks: decision-making without empathy, misalignment with brand tone, and experiences that feel generic or alienating.

But no ecommerce leader wants to enter the “uncanny valley” territory that artificial experience can create — where interactions feel almost human but lack genuine understanding or empathy. This manifests in several critical ways that can damage brand relationships and reduce conversion rates . According to a recent report published by MIT, 95% of generative AI pilots are failing to deliver business value. At Infosys Equinox, we believe the missing link is human-centricity.

• Robotic conversational experiences: Many AI chatbots and shopping assistants fall into scripted responses that feel mechanical and disconnected from customer emotions. When a frustrated customer explains a complex return issue, an AI that responds with generic troubleshooting steps instead of acknowledging their frustration creates a poor experience that drives customers away.
• Over-personalization without context: AI merchandising systems can become intrusive when they lack proper contextual understanding. Recommending baby products to someone who recently experienced a miscarriage or continuing to promote vacation packages after a customer has experienced a family emergency demonstrates how data-driven personalization can feel tone-deaf without human sensibility.
• Lack of Emotional Intelligence: AI systems that cannot recognize emotional cues — whether in text, voice, or behavioral patterns — often escalate situations unnecessarily. A customer expressing urgency about a delayed shipment for a wedding gift needs immediate escalation, not a standard "we'll get back to you within 24 hours" response.
• Inconsistent Brand Voice: When AI systems aren't properly trained on brand values and communication style, they can deliver messages that feel disconnected from the brand's personality. A luxury brand's AI assistant that uses casual language or a family-friendly retailer's AI that fails to maintain warmth can undermine brand positioning.

Redefining Human-Centricity for the AI Age

Design for Emotional Intelligence -Human-centric commerce is emotionally textured, full of delight, frustration, urgency, and surprise. If AI systems lack emotional sensitivity, they risk breaking rapport. For instance:

Here are five core principles for maintaining human-centricity in the age of agentic AI:

1. Serve Human Needs, Not Just Business Metrics

To stay human-centric, brands must embed empathy into the objective function. That means:

• Designing AI agents that prioritize customer goals alongside revenue (e.g., minimizing returns, reducing choice overload).
• Using behavioral signals (hovering, hesitating, refining search) as signs of unmet needs, not conversion failure.
• Rewarding long-term trust and loyalty over short-term gains.

Agentic systems should act like helpful concierges, not aggressive salespeople.

2. Maintain Transparency and Control

One of the paradoxes of AI is that it offers more power while creating less visibility. When a shopping assistant recommends a product or a bot makes a substitution in a subscription box, the reasoning can feel opaque.

Human-centric commerce demands transparency. Customers should:

• Understand when they're interacting with an AI vs. a human.
• Be able to inspect, override, or decline AI decisions (e.g., "Why was this recommended to me?" or "Don’t show this again").
• Receive plain-language explanations of complex logic (e.g., “We picked this because it matches your past purchases and is trending among people like you”).

This builds trust and ensures that AI remains an augmenter, not a dictator.

3. Design for Emotional Intelligence

Human-centric commerce is emotionally textured — full of delight, frustration, urgency, and surprise. If AI systems lack emotional sensitivity, they risk breaking rapport. For instance:

• A chatbot that responds to a complaint with tone-deaf cheerfulness.
• A shopping assistant that recommends high-priced luxury items during a customer’s budgeting phase.
• A personalization engine that shows pregnancy products to someone recovering from a loss.

To avoid this, AI systems must be trained not only on data but on emotional context. This can include:

• Using tone detection and sentiment analysis to adjust responses.
• Designing fallback states that acknowledge limitations (e.g., “I’m not sure I understood that — want to rephrase or speak to a human?”).
• Building personas that reflect the brand’s empathy and tone-of-voice, not just its product catalog.

4. Implement Seamless Human Handoffs

The most effective AI commerce systems recognize when human intervention is needed and facilitate smooth transitions to human agents. This requires sophisticated intent recognition and the ability to pass complete context to human representatives, ensuring customers don't have to repeat information or start over.

5. Respect the New Agency of the Customer

In an agentic AI future, humans won’t just receive help from bots — they’ll send bots to do their bidding. Shopping agents like Apple’s Siri, Amazon’s Alexa, or OpenAI’s ChatGPT may soon act on a customer’s behalf across multiple sites, filtering noise, comparing offers, and placing orders based on known preferences.

This shift changes the audience of your digital experience. You’re no longer just designing for humans but also for their agents, which means:

• Exposing machine-readable data like specs, reviews, and inventory cleanly and consistently.
• Offering APIs and feed structures that allow agents to negotiate, customize, and compare.
• Ensuring that the “first impression” your site gives to autonomous agents is rich, structured, and aligned with your human experience.

This isn’t just a technical shift — it’s a philosophical one. Are you treating AI agents as adversaries to outwit, or as proxies for your customers to collaborate with?

Use Cases That Center the Human in the Loop

Let’s explore a few real-world examples of how brands can embed human-centricity even within deeply automated experiences:

Shopping Assistant on a Beauty Site

• Human-centric approach: The assistant asks preference questions in a friendly, non-invasive way (e.g., “Are you shopping for yourself or a gift?”), and recommends based on skin tone, concern, and budget.
• Agentic layer: It tracks usage across sessions, notices if a user prefers fragrance-free products and adapts suggestions accordingly — not just once, but over time.

AI Merchandiser for Fashion Ecommerce

• Human-centric approach: Search results adapt to intent (e.g., “red dress for wedding” brings festive items, not office wear) and editorial content is inserted to inspire or guide.
• Agentic layer: The engine re-ranks products based on return likelihood and matching size availability, reducing the emotional pain of failed fits or unavailable items.

AI Customer Service Chatbot

• Human-centric approach: The bot doesn’t just answer questions, it detects when a customer is overwhelmed, offers to simplify information and routes to a human when empathy is needed.
• Agentic layer: The bot learns from past resolutions, proactively offers status updates and nudges customers toward quicker resolutions — all while maintaining tone consistency.

Metrics for Measuring Human-Centric Automation

To ensure you're not just deploying automation for automation’s sake, consider tracking metrics that reflect human outcomes:

• Customer Frustration Rate: How often does a user abandon chat, repeat queries, or escalate to human help?
• Emotional Sentiment Scores: How does the customer's tone evolve throughout an automated interaction?
• Agent Override Ratio: How often do customers dismiss or undo an AI recommendation?
• Experience Net Promoter Score (xNPS): Would the user recommend the AI-driven experience itself?

These metrics prioritize not just speed or conversion, but satisfaction, confidence, and emotional resonance.

Building Trust Through Human-Centric AI

The age of agentic AI in commerce is not about replacing human experiences, it's about enhancing them. By implementing AI systems that prioritize empathy, context and authentic human connection, enterprise commerce leaders can serve automated experiences that feel genuinely human while delivering the efficiency and personalization that modern customers expect.

At Infosys Equinox, we believe that human-centricity must remain at the heart of digital commerce, even as AI takes on a more autonomous role. Our platform is built to ensure that agentic AI enhances, not erodes, the emotional and contextual richness of customer interactions. With a composable, API-first architecture and advanced agentic capabilities, our platform empowers brands to design intelligent systems that collaborate with customers and their agents - while preserving brand voice, context, and trust.

Whether you're optimizing customer experience, empowering business users, or automating your product development, Infosys Equinox helps you build a network of AI agents that go far beyond simple automation benefits by building deeper customer relationships and more resilient business models in an increasingly AI-driven world.

Ready to lead with human-centricity in an AI-driven world? Let Infosys Equinox help you build the future of commerce. Write to us at contactus@infosysequinox.com.