top of page
Search

Navigating the Future of AI: Insights from My Journey as a Product Manager in the Agentic AI world

Artificial intelligence is reshaping how we live and work, and as someone deeply involved in that shift, I want to share what I’m learning along the way. As a product manager working in an era where customers have an ever-growing appetite for AI solutions, I get a close view of both the opportunities and the hard parts. Outside my day job I also build agentic AI systems hands-on — including a self-running pipeline that tracks the AI landscape from end to end — so this is written from the perspective of someone shipping these things, not just watching them. Here I’ll document my journey, share the projects I’m working on, and explore how AI is influencing product management, the IT industry, and society at large.



My Journey into Agentic AI


Agentic AI refers to systems that can act autonomously to achieve goals, make decisions, and adapt to changing conditions. The potential is vast — and so are the unknowns. My role as a product manager is to guide AI products that are not only capable, but ethical, user-friendly, and genuinely useful.


What my projects have taught me most is the importance of balancing autonomy with control. Building my AI-news pipeline, for example, meant letting the system collect, de-duplicate, and summarize on its own — while keeping firm guardrails around what it is allowed to do and clear checkpoints where I stay in the loop. That balance, autonomy inside well-defined limits, is what keeps an agent aligned with user expectations and business goals.



Insights from AI Projects in Product Management


Managing AI projects calls for a different approach than traditional software development. A few lessons that have stuck with me:


  • Clear Objectives

Defining exactly what the AI should achieve, up front, is crucial. Ambiguous goals lead to wasted effort and murky outcomes.


  • Data Quality Matters

AI systems live and die by their data. Keeping data sets clean, relevant, and unbiased is a continuous effort.


  • Reliability and Evaluation

Agents fail in new ways — they act, not just answer — so reliability has to be designed in, with evaluation, observability, guardrails, and an eye on running cost from day one, rather than bolted on later.


  • User-Centered Design

Even autonomous AI exists to serve people. Involving users early shapes features that genuinely add value.


  • Ethical Considerations

AI touches privacy, fairness, and transparency, so ethical guidelines belong in the development process from the start.


AI decision-making can be unintentionally biased when training data is skewed. Fixing it may mean revisiting data sources and retraining models — which can slow timelines, but makes the product fairer and more trustworthy.



The Latest AI News and Its Impact on Product Management


The story has shifted from AI that talks to AI that acts. The frontier now is agentic AI — systems that plan, use tools, and complete multi-step work with limited supervision — alongside a regulatory regime that is finally in force. A few developments shaping product decisions:


  • From chatbots to agents

Models now reason and act, not just answer, and AI is moving from a feature inside an app toward becoming the interface itself. Gartner expects 40% of enterprise applications to embed task-specific AI agents by the end of 2026, up from less than 5% a year earlier.


  • The demo-to-production gap is real

Many companies say they have adopted agents, but far fewer run them reliably in production, and Gartner research suggests more than 40% of agentic projects risk cancellation by 2027 over cost, unclear value, and weak governance. The lesson for product managers: aim agents at narrow, well-defined workflows, and treat reliability, cost, and governance as product requirements from day one.


  • Regulation is binding, not hypothetical

Under the EU AI Act, bans on the riskiest practices (since February 2025) and general-purpose AI rules (since August 2025) already apply; transparency and content-labelling rules arrive in August 2026, while the heaviest high-risk obligations have been pushed to December 2027. Fines reach €35 million or 7% of global turnover, and the rules apply to anyone whose AI touches people in the EU. If your product screens candidates or scores credit, it likely counts as high-risk — so classify it early.


These trends reward product managers who stay agile: build governance in early, keep learning, and be ready to pivot as both capabilities and rules evolve.



Eye-level view of a futuristic AI control panel with dynamic data visualizations
AI control panel showing real-time data and decision pathways


AI’s Effects on the IT Industry and Beyond


AI is reshaping how software is built, tested, and maintained. The shift goes beyond autocomplete: AI coding agents now help write, test, and review code, and agentic tooling increasingly handles monitoring, security, and routine operations — compressing development cycles and changing what engineering teams spend their time on.


Beyond IT, AI is making its mark across sectors:


  • Healthcare

Assisting with diagnostics, personalized treatment, and drug discovery.


  • Finance

Strengthening fraud detection and risk assessment through pattern recognition.


  • Manufacturing

Reducing downtime with predictive maintenance.


  • Education

Adapting learning experiences to individual students.


It is worth noting that several of these — healthcare, finance, and hiring among them — are exactly the areas the EU AI Act treats as high-risk, so innovation in these fields now comes with real compliance obligations. More broadly, as AI takes over repetitive work, the human premium shifts toward oversight, creativity, and complex problem-solving.



Looking Ahead: The Role of Product Managers in an AI-Driven World


Product managers will help shape where AI goes next. To do that well, we need to:


  • Champion User Needs

Ensure AI products solve real problems and enhance user experiences.


  • Promote Transparency

Help users understand how AI makes decisions.


  • Own Accountability

Take responsibility for governance — what an agent is allowed to do, where humans stay in the loop, and how outcomes are audited.


  • Balance Innovation and Responsibility

Push boundaries while safeguarding ethical standards.


  • Foster Collaboration

Work closely with engineers, designers, ethicists, and users.


By embracing these responsibilities, product managers can steer AI toward outcomes that are good for businesses and for society.



Sources & Further Reading

•   AI Agent Trends 2026 Report — Google Cloud, 2026

•   Regulatory Framework for AI (the AI Act) — European Commission, updated June 2026



 
 
 

Comments

bottom of page