{ design@tive } information design

Itamar Medeiros, from Brazil, currently lives in Bonn (Germany), where he works as Senior User Experience Designer for Roamworks. Working in the Information Technology industry since 1998, Itamar Medeiros has developed several websites and interactive media campaigns, both as art director and, later, as creative director for several new media agencies in Pernambuco, Brazil. As an Information Design specialist, served as board member of the Brazilian Society of Information Design, lead the Design Research team of the Hypermedia Lab at the Universidade Federal de Pernambuco, and held a lecturing position as Assistant Program Director of the Visual Communication Department of the Raffles Design Institute at DongHua University, China. Recently, he has been awarded with an MA (distinction) in Design Practice at Northumbria University (Newcastle, UK) for his thesis “Designing the User Experience: Creating Innovative Design Software Solutions within Collaborative/Distributed Design Environments”

[designative]

Is belief part of being Brazilian, or just how we set ourselves up for heartbreak? Tonight, Brazil plays Japan. By the end, Brahma’s “Tá Liberado Acreditar” will feel either painfully naive — or exactly right. #advertising #football #inspiration

For Inspiration: "Tá Liberado Acreditar"

[designative]

Users don't experience AI models. They experience the harness around them. Context, memory, tools, planning, verification, and governance determine whether AI becomes reliable, governable, and trustworthy in practice. #AIHarnessEngineering #ArtificialIntelligence #HumanAgentCenteredDesign

What Designers Need to Know About AI Harness Engineering

[designative]

The greatest risk in agentic systems isn't that humans disappear. It's that they remain involved while nobody can clearly explain who was accountable for the outcome. #ArtificialIntelligence #AgenticAI #AIEthics #HumanAgentCenteredDesign

Make Human Accountability Traceable: A Primer of Human-Agent Interaction Guidelines

[designative]

What if the biggest coordination risk in agentic systems isn't too little information, but too much? As agents summarize, prioritize, and mediate collaboration, how do teams maintain common ground instead of drifting into fragmented understanding? #AgenticAI #ArtificialIntelligence

View On WordPress

[designative]

We keep asking whether #AI can explain itself. But in agentic systems, the bigger challenge may be visibility. The issue isn't understanding decisions after they happen—it's knowing what the system is doing while work is still unfolding. Transparency matters, but awareness enables trust. #AgenticAI

Expose System State: A Primer of Human-Agent Interaction Guidelines

[designative]

#Accessibility failures in agentic systems do not always look like exclusion. More often, they look like missing disagreement, silent participants, incomplete organizational memory, or teams mistaking workflow silence for real alignment. #AgenticAI #ArtificialIntelligence #AIUX #UXforAI #Usability

Ensure Inclusivity and Accessibility: A Primer of Human-Agent Interaction Guidelines

[designative]

Most #AI #UX discussions still treat feedback as a #usability feature. But once agents help prioritize, synthesize, and coordinate work, feedback becomes governance infrastructure. The challenge is designing adaptive systems. that remain governable without turning people into full-time supervisors.

Embed Feedback Loops in the UI: A Primer of Human-Agent Interaction Guidelines

[designative]

We often discuss “trustworthy AI” as if trust is the goal. But what if systems become reliable before humans can properly inspect or intervene? The challenge may not be trust, but calibrated reliance. #AgenticAI #AI #TrustBuilding #HumanAgentCenteredDesign #TrustCalibration

Trust Calibration in Agentic AI: Designing for Appropriate Reliance, Not Blind Trust

[designative]

Most organizations think they preserve human control because someone still approves the #workflow. Meanwhile, #orchestration systems already shape what people see, prioritize, escalate, or ignore. This article argues that control is about governable delegation — not approval screens. #AgenticAI

Preserve Human Control: A Primer of Human-Agent Interaction Guidelines

[designative]

Most #AI UX still treats conversation as #interface polish. But to shape dialogue patterns in agentic systems is really to design #coordination, #escalation, #oversight, and organizational #trust. What happens when dialogue becomes operational infrastructure? #AgenticAI #HumanAgentCenteredDesign

Shape Dialogue Patterns: A Primer of Human-Agent Interaction Guidelines

[designative]

Most #AI transparency discussions assume a single model explaining a single output. But what if #reasoning involves multiple layers, workflows, and #agents? The challenge might not be explainability but interrogability. #AgenticAI #ArtificialIntelligence #InteractionDesign #HumanAgentCenteredDesign

Make Reasoning Interrogable: A Primer of Human-Agent Interaction Guidelines

[designative]

What if the most important product decisions in #AI systems no longer live in #interfaces, #PRDs, or #journeymaps? They now live inside routing logic, memory systems, escalation policies, and #orchestration pipelines. Are product teams reviewing those layers? #AgenticAI #ArtificialIntelligence

What Designers Need to Know About AI Orchestration

[designative]

Engineers run the AI evals. But who decides what “good” actually means? If your criteria only measure what’s easy, your product will optimize for the wrong things. Should designers and PMs own eval criteria? Let’s debate. #AIEvals #ProductDesign #UX #AgenticAI #DesignLeadership #UXforAI #AIUX

AI Evals for Designers and Product Managers: Why Criteria Matter Most

[designative]

We talk a lot about what AI agents should do. But who defines what they’re allowed to do? Workflow encapsulation shifts design from interfaces to boundaries and makes design leadership a governance issue. Are you ready for that shift? #DesignLeadership #AgenticAI #UXStrategy #DesignManagement #AI

Designing Boundaries, Not Features: Workflow Encapsulation as a Leadership Discipline

[designative]

#AgenticAI breaks traditional requirements. If systems interpret goals, are outcomes enough—or do we need explicit intent? Intent documentation in agentic #ProductDesign may be the missing layer. Curious how teams should adapt? #ArtificialIntelligence #DesignManagement #DesignStrategy

From Intent to Alignment: Why Design Leadership Must Institutionalize Intent Documentation in Agentic Product Design

[designative]

We can trace everything our AI systems do—but can we tell if it’s actually good? Evaluation in agentic AI is not about more metrics. It’s about defining what “good” means—and testing it continuously. Are your #evals designed—or just observed? #AgenticAI #ProductStrategy #UXStrategy #ProductDesign

From Behavior to Judgment: Designing Evaluation for Agentic Systems

[designative]

We’ve focused on explainability—but is that enough? If you can explain a decision but can’t see how the system behaves over time, can you really trust it? Maybe the real gap in AI design isn’t reasoning… it’s observability. #AI #UX #ProductStrategy #HumanAgentInteraction #Observability #Metrics

AI Observability Is the Missing Layer in Human–Agent Systems