Tomas Gear leads Parloa’s Agent Engineering team in New York, where he architects and launches AI voice agents that power customer conversations for Fortune 500 enterprises. As Parloa’s first U.S. technical hire, he has built the North-American delivery organization and has since overseen some of the sector’s largest, most complex deployments. Tomas is regarded for his deep expertise in conversational-agent engineering, orchestration, enterprise-grade integrations, and advanced prompting strategies.Before joining Parloa, Tomas held software and data-engineering roles at multiple high-growth startups, designing large-scale, data-driven systems that accelerated product innovation. He is a thought leader on applied and voice AI, agent engineering, and the future of customer experience.
Artificial intelligence has already radically changed how companies can serve their customers, and those changes will only accelerate. Agentic AI is unlocking new possibilities, but it’s also prompting business leaders to look ahead and prepare to harness all this new potential.
Read on to peer into the next one, three, and five years of agentic AI, and the organizational adaptations, technological adaptation, and ethical frameworks needed to make customer interactions smarter and more efficient—but also more human.
Year One: Conversations and Connection Over Call Resolution
Today’s customer contact centers aim to make responses and resolutions as rapid as possible. Fast enough isn’t fast enough. The lower the average handle times, the better. But in the year of the proliferation of agentic AI, that paradigm is going to flip.
AI is rapidly moving beyond basic chatbot and FAQ functionality, handling routine inquiries and requests with increasing accuracy, empathy, and personalization. More and more, rather than deflecting customers with canned responses, these AI agents will engage in genuine conversations, seeking to understand customers on a deeper level.
Agentic AI won’t just resolve customer requests—it will help build relationships. To do that, it will personalize interactions by gathering and responding to contextual information through sustained interactions.
To leverage this ability, businesses will need to not only make technological changes, but cultural changes as well. They must optimize less for transactional KPIs and more for customer trust, satisfaction, and lifetime value.
Meanwhile, support teams will need to embrace agentic AI not as their replacement, but as a partner in their work. Human agents will still be essential, but their roles will evolve toward solving more complex problems, offering empathy to customers in certain cases, and guiding strategy.
Year Three: One-to-One AI Agents for Every Customer and AI-to-AI Commerce
At the three-year mark, we’ll look back on traditional websites and apps as unwieldy relics of a bygone internet age.
By 2028, conversational AI will maintain rich and continuous context across channels, including voice, chat, texting, email, etc. Every customer will have their own AI agent that will learn and remember their preferences, purchasing patterns, and support history. Just as importantly, they’ll constantly use that information to truly personalize every interaction. Zendesk has already predicted that by 2030, AI will play a role in 100% of all customer interactions.
At the same time, the way customers initiate those interactions will fundamentally change. Rather than searching, clicking, or tapping their way to a brand’s website or app, customers will simply ask their personal AI assistant to do their shopping for them. Whether it’s booking or changing a flight, scheduling or canceling a service, or ordering a product, the customer will leave the work to their AI agent, who will coordinate with each brand’s agentic AI.
Key to success at this point will be setting up and refining the infrastructure to support these AI-to-AI interactions. Functionally, companies will need to have an AI agent for every customer (and their AI agent). The monolithic chatbot will become a thing of the past, replaced by millions of individualized AI customer service representatives, each making full use of personalized data to make interaction efficient, effective, and satisfying.
To achieve this, businesses will need to be capable of supporting high volumes of concurrent and constantly personalized interactions. To make that feasible, they’ll need to ensure interoperability with all major assistant platforms, and to integrate with predictive analytics, equipping their AI agents to anticipate—not just react to—customer requests, questions, and complaints.
The scale of this idea might seem unlikely, but in fact we’re already moving in this direction. Companies that want to get ahead or stay ahead should start preparing sooner than later.
Year Five: The End of Websites
By 2030, most customers will never navigate menus or search results. Instead, they’ll simply express a need through voice, text, or gesture, and their AI assistant will collaborate with brand-side agents to fulfill that need with incredible speed and specificity. Booking a vacation, troubleshooting a device, subscribing to a service—all this will be orchestrated by autonomous AI on both sides of the interaction.
These AI agents will become increasingly predictive, learning from each interaction and blending emotional intelligence and memory to help customers proactively and make increasingly effective real-time decisions.
While this point is all but inevitable, there is still much to be done to get there. We’ll need to radically advance natural language understanding, increasing the capability to interpret nuanced multimodal communication.
Ethical frameworks will need to be in place and robust enough to protect customer privacy, ensure transparency, and prevent unfair practices. Data integration will need to be highly secure and efficient, so that customers are protected without impairing the functionality of AI agents.
And—perhaps most importantly—systems and talent for human oversight will be needed to continually optimize functionality and guide ongoing strategy. In other words, humans should not be removed from this system, but AI should be used to elevate their contributions, maximizing their potential.
We are already seeing the beginnings of this with the introduction of AI-driven web browsers such as Perplexity’s Dia, which features an embedded chatbot. This highly intelligent chatbot can quickly and accurately summarize materials, answer questions, and feed information to users within the web browser, so they never need to leave the browser and visit a website.
Final Thoughts: The Bridge to the Future Is Built Through Ecosystems
In one year, we’ll see AI as a trusted assistant.
In three, as a personalized agent working across systems and platforms.
In five, as the very fabric of interaction — invisible, intelligent, indispensable.
Fully realizing these possibilities will require vision, investment, and a willingness to rethink how we define customer experience. No one vendor will get there alone. As Forrester states, “B2B organizations are increasingly relying on partner ecosystems to fulfill buyer and customer expectations, advance innovation opportunities, and achieve corporate revenue and growth objectives.”
Leading companies like Salesforce, ServiceNow, and Microsoft are rethinking their partner relationships, viewing them as more than just distributors or resellers. Instead, these stakeholders are coming together with the goal of joint value creation, integrating their technologies to solve challenges and deliver the best customer experience across any channel, based on customer preference.
This won’t be only about one silver bullet technology, but about many vendors coming together collaboratively, using technology to make customer interactions more meaningful, productive, and empathetic.
