AI Agents and the New Era of Artificial Intelligence

In recent years, artificial intelligence has been advancing by leaps and bounds. First, we automated repetitive tasks; then we moved on to generating content with AI; and today, we have entered the era of autonomous agents—systems capable of thinking, planning, and executing entire processes with minimal human intervention.

AI Agents vs. Generative AI

Between 2005 and 2020, automation moved data and connected applications based on fixed rules. With generative AI, tools like GPT-3 and ChatGPT could automatically create content but could not act upon it. AI agents change that. They combine language models, persistent memory, external tools, and feedback loops, enabling the execution of complete processes autonomously.

Market and Opportunities

The global market for AI agents could reach $93–$196 billion between 2032 and 2034, with annual growth exceeding 40%. In Europe, investment in agent-focused startups grew 150% in 2025.

While major players focus on standardized platforms, vertical startups that automate critical processes (finance, healthcare, logistics) gain sustainable advantages through sector expertise, proprietary data, and regulatory barriers.

For example, one company in our portfolio, Metly:

In the pharmaceutical sector, its agents analyze market data, prioritize opportunities, and coordinate manufacturing processes and partnerships.
Result: less manual intervention, faster processes, and more precise strategic analysis.

How AI Agents Work

At their core, an autonomous agent combines:

A “brain” that plans,
Memory to retain information,
Tools to execute tasks,
And a control system that intervenes only when necessary.

They can take on different roles:

Researchers and analysts: prioritize issues and generate insights.
Planners: define strategies and action sequences.
Executors and communicators: act on systems and clients.
Coordinators and evaluators: supervise workflows and validate results.
Learners and creatives: adjust behavior and generate reports.

Why Invest in AI Agents

The value lies not only in the technology but also in its integration into real processes and sector specialization. Business models are evolving: some charge per use, others per outcome, and many combine subscription, consumption, and results-based fees.

At Encomenda, we do not invest in “AI for AI’s sake.” We look for teams that apply agents practically to solve critical problems.

For example, one of our portfolio companies, Uxia:

Automates user testing by generating synthetic profiles, prioritizing scenarios, and validating product experiences in minutes.
Benefit: accelerates complex processes, reduces costs, and allows faster iteration with tangible results.

In short, autonomous agents are already here. Understanding their technology, integration, and sector specialization will allow companies and investors to capture unique opportunities in the emerging agent-driven economy.

Watch the full MasterClass by Mario Brassesco from Encomenda here: https://youtu.be/_3FElHUgkoA