Human-Centered AI Systems Designer

Human-centered AIsystems forreal operationalwork.

I help teams simplify communication, workflows, and decision-making through responsible AI, design, and systems thinking.

Turning complexity into clear, useful systems.

I design human-centered AI systems that simplify operations, communication, and complex workflows.

Visual Clarity

Brand Identity & Visual Systems

Premium, flexible identity systems that make a brand easier to recognize, explain, and use across touchpoints.

Human AI

AI Workflow Design

Responsible AI flows that turn rough inputs into useful drafts, summaries, decisions, and documents without hiding the human judgment layer.

Calm Operations

Automation & Productivity Systems

Lean internal systems that reduce repeated work, organize information, and make everyday processes easier to follow.

Teams and organizations dealing with operational overwhelm.

The best fit is not “people who want AI.” It is people who have communication overload, unclear workflows, repeated decisions, and systems that need to become easier to understand.

Operations-heavy teamsGroups managing repeated work, internal requests, documentation, and process handoffs.
Public service and frontline service environmentsTeams balancing policy, people, service delivery, and communication under pressure.
Small businesses and foundersOwners and leaders overwhelmed by tools, decisions, documents, and scattered operating knowledge.
Communication-heavy teamsPeople who spend too much time rewriting, clarifying, following up, and explaining the same things.
Responsible AI explorersOrganizations that want AI to support humans, not replace judgment or create new risk.

A practical process for designing AI systems people can actually use.

The work starts with the human workflow, not the tool. AI only enters after the friction is understood and the system has a clear purpose.

Understand the workflowMap what people are trying to do and where the work gets stuck.
Identify frictionFind repeated decisions, unclear handoffs, missing context, and communication drag.
Simplify the systemClarify language, inputs, outputs, and the decisions the system needs to support.
Design the AI-supported workflowCreate the structure, templates, workflow, prompts, and review points.
Keep humans in reviewUse AI where it reduces friction without removing human control.
Test and improveCheck whether the system works under real pressure, then refine it through use.

The work is designed to be seen, tested, and improved.

Every case study is moving toward visual proof: before and after workflow maps, input-to-output diagrams, system diagrams, prompt architecture, and process maps that make AI systems design easier to evaluate.

Before → After WorkflowShows how scattered work becomes a simpler operating path.
Input → AI Support → Human Review → Final OutputMakes responsible AI visible instead of hidden.
System DiagramsClarifies roles, steps, decision points, and feedback loops.
Prompt ArchitectureDocuments how instructions, guardrails, and review criteria shape output.
Process MapsTurns operational knowledge into something teams can inspect and reuse.

Responsible AI should make work clearer and people more capable.

The work avoids generic AI hype. The goal is to design systems that reduce stress, protect judgment, and help people understand what is happening.

AI should support humans, not replace themThe system helps with structure, drafting, summary, and review without removing accountability.
Clarity over complexityHuman-centered AI should make communication workflows easier to understand.
Human review mattersPeople need visible checkpoints before a draft, decision, or response becomes final.
Privacy and responsibility come firstResponsible AI means naming risk, protecting sensitive details, and avoiding invented facts.
Systems should reduce stressGood workflow automation removes confusion instead of creating another layer to manage.

Work that turns messy moments into clear systems.

Each project turns a messy moment into a clearer path: what to say, what to do next, where the source lives, or how a process should run.

Email + Docs Drafting
Claude / Docs

AI Email & Docs Specialist

A drafting workflow for turning rough notes and voice thoughts into clear, polished communication.

ProblemDrafting takes too long when ideas are scattered.
SystemAI-assisted writing and review flow.
OutcomeClearer drafts, faster review, and visible human approval.
View full case study
SOP + Process Map
ChatGPT / Mapping

SOP Generator + Process Mapper

A practical system that turns raw notes into usable SOPs, checklists, and visual process maps.

ProblemProcess knowledge lives in heads, chats, and fragments.
SystemDocumentation and process design workflow.
OutcomeReusable SOPs, clearer handoffs, and process maps teams can inspect.
View full case study
Private Care Service
Brand / Workflow

Private Nurse

A focused service concept for making private care feel more understandable, warm, and coordinated.

ProblemCare choices can feel stressful, unclear, and fragmented.
SystemBrand and service clarity system.
OutcomeWarmer trust, clearer intake, and a calmer path for families.
View full case study
Task + Knowledge Hub
Notion / AI

ClearPath OS Lite

A lightweight workspace for organizing tasks, decisions, notes, and repeatable workflows in one calm place.

ProblemDaily work gets spread across too many tools and half-systems.
SystemProductivity and automation workspace.
OutcomeLess scattered work and a stronger foundation for responsible automation.
View full case study
Public Operations AI
Concept / Operations

Human-Centered AI for Public Operations

A conceptual system for helping frontline teams reduce communication overload, clarify policy, and keep humans in review.

ProblemComplex service environments create stress, delays, and inconsistent communication.
SystemResponsible AI workflow for triage, plain-language summaries, and human review.
OutcomeClearer public service operations without removing judgment from people.
View full case study

Technology should not make people feel smaller. Good design and responsible AI should reduce confusion, protect human judgment, and make everyday work easier to understand.

Fabian Delven designs systems that feel calm enough to use.

I work between graphic design, automation, and AI implementation. The goal is not to make tools look impressive for a demo. The goal is to make decisions, workflows, and communication easier for real people doing real work.

More About Me

Have a workflow, idea, or system that feels messy? Let's make it clear.

Bring the rough version: the scattered notes, the repeated process, the fuzzy idea, the tool stack that almost works. I will help shape it into something useful.