Topics In Demand
Notification
New

No notification found.

Welcome to the Age of AI Agents: Q&A with Vitor Domingos, Principal Architect, EMEA, Hitachi Digital Services
Welcome to the Age of AI Agents: Q&A with Vitor Domingos, Principal Architect, EMEA, Hitachi Digital Services

March 4, 2025

20

0

Welcome to the age of AI Agents! Can you explain what these AI Agents are and how they’re transforming the workplace?

Absolutely. We’re entering a transformative era where AI Agents are becoming much more than tools - they’re poised to operate as digital colleagues. Think of them as highly capable team members, able to manage workflows, enhance productivity, and assist with decision-making. Unlike traditional chatbots, which are largely reactive, today’s AI Agents are evolving into digital twins. These systems are autonomous, capable of performing complex tasks, and even making decisions based on high-level instructions.

Imagine a project manager powered by AI that can plan, sequence, and execute tasks, or a virtual assistant capable of collaborating across departments. This is no longer science fiction - it’s rapidly becoming a reality. However, it’s essential to approach this evolution with caution and responsibility, especially regarding security, transparency, and ethical compliance.

You emphasize that trust is key for AI Agents to succeed. Why is this so critical, and what challenges are organizations facing in building trustworthy AI systems?

Trust is the cornerstone of any technology adoption, and AI Agents are no exception. For organizations to rely on these systems, AI Agents must operate transparently and align with privacy laws and responsible standards. Without trust, they’ll remain a novelty rather than indispensable tools.

The challenges include reliability, governance, and explainability. For example, AI Agents sometimes “hallucinate” or generate incorrect outputs, which can be catastrophic in high-stakes scenarios. Furthermore, integrating AI Agents into existing workflows often requires substantial investment in infrastructure and talent - both of which can be in short supply.

Organizations also need to prioritize explainability. Stakeholders must understand how and why an AI Agent makes certain decisions, particularly in regulated industries like fintech or healthcare. Without this, building trust and ensuring compliance with regulations like GDPR or PCI DSS becomes a daunting task.

Let’s talk about the anatomy of AI agents. Could you walk us through these?

Certainly. AI Agents are underpinned by four essential components:

                  1. Large Language Models (LLMs): These are the brain of the system, interpreting user instructions and generating action plans. They’ve become incredibly sophisticated since the release of ChatGPT in 2022, enabling more nuanced and capable interactions.

                  2. Tools: These extend the agent’s functionality, enabling it to perform tasks like web searches, document retrieval, and data visualization. This modularity makes AI agents highly versatile.

                  3. Memory: Memory is what allows agents to handle complex, multi-step tasks. It combines long-term access to databases with short-term memory to manage contextual information.

                  4. Reflection and Self-Critique: This is the cutting-edge capability that allows AI agents to evaluate their performance in real time, identify errors, and refine their strategies. It’s akin to having a built-in quality assurance mechanism.

Together, these components form the backbone of agentic systems, enabling them to operate autonomously and collaboratively.

What’s driving the rapid evolution of these systems, and what potential barriers still exist?

The evolution of AI Agents has been fuelled largely by advancements in LLMs and the increasing demand for intelligent automation. We have seen early AI copilots, limited to one-off tasks, but today’s systems are far more capable, functioning like digital project managers or even collaborators.

However, some barriers remain. And these include:

                  • Cost and Complexity: Building and deploying these systems requires substantial resources.

                  • Talent Shortages: There’s a significant gap in AI expertise, which slows development.

                  • Reliability Issues: AI systems can still make errors or produce nonsensical results.

                  • Integration Challenges: Legacy systems can be difficult to align with cutting-edge AI solutions.

Despite all these issues, the potential is immense. Strategic investment and planning are key to overcoming these barriers. So, organizations need to think ahead.

You also highlight the importance of responsible AI. How should organizations approach the design of trustworthy systems?

It’s not easy for most, but trustworthy AI Agents must meet several key criteria:

  • Explainability: Stakeholders should be able to trace and understand the decisions made by AI agents.
  • Bias Mitigation: Agents must be designed to detect and reduce biases, ensuring fair and equitable outcomes.
  • Sustainability: Energy-efficient operations are increasingly important, both for cost savings and environmental impact.
  • Security: Robust safeguards against adversarial attacks are essential to maintaining reliability.

Embedding these principles into the design process, is key, and then organizations can build AI Agents that are not only effective but also aligned with societal and regulatory expectations.

Finally, where do you see AI Agents heading in the near future?

Good question! AI Agents are evolving from tools to dynamic collaborators, as I said, but we’re already seeing systems that can adapt to new information, collaborate with other agents, and tackle increasingly complex tasks. The next phase will likely involve greater specialization, where agents focus on specific industries or roles.

For CIOs and CTOs, the message is clear: the age of AI Agents is here, and the opportunities are enormous and open to exploration. However, success requires more than technology - it demands governance, explainability, and a clear alignment with organizational goals. The road ahead is challenging, but the organizations that lead the charge will gain a decisive edge in a world increasingly driven by intelligent automation.


That the contents of third-party articles/blogs published here on the website, and the interpretation of all information in the article/blogs such as data, maps, numbers, opinions etc. displayed in the article/blogs and views or the opinions expressed within the content are solely of the author's; and do not reflect the opinions and beliefs of NASSCOM or its affiliates in any manner. NASSCOM does not take any liability w.r.t. content in any manner and will not be liable in any manner whatsoever for any kind of liability arising out of any act, error or omission. The contents of third-party article/blogs published, are provided solely as convenience; and the presence of these articles/blogs should not, under any circumstances, be considered as an endorsement of the contents by NASSCOM in any manner; and if you chose to access these articles/blogs , you do so at your own risk.


Hitachi Digital Services, a wholly owned subsidiary of Hitachi Ltd., is an edge-to-core digital consultancy and technology services provider helping organizations realize the full potential of AI-driven digital transformation. Through a technology-unified operating model for cloud, data and IoT, Hitachi Digital Services' end-to-end value creation for clients is established through innovation in digital engineering, implementation services, products, and solutions. Built on Hitachi Group's more than 110 years of innovation across industries, Hitachi Digital Services helps to improve people's lives today and build a sustainable society tomorrow. To learn more, visit https://hitachids.com.

© Copyright nasscom. All Rights Reserved.