That’s how trust is built. That’s how intelligence is shaped.
And that’s what I design.
I design AI systems by shaping how they reason, respond, and earn trust in practice.
Trusted by teams at:
I help teams design, build, and evolve AI systems that drive efficiency, consistency, and customer trust.
Conversation Design
Crafting natural, context-rich dialogues that help users feel understood and supported.
Consulting
I advise organizations on AI systems design, evaluation strategy, and feedback architecture. Whether you’re building from scratch or scaling an existing platform, I help connect intent, performance, and impact.
AI Evaluation Systems
I turn metrics into insight, spotlight optimization opportunities, and guide smarter model improvement.
Workshops & Trainings
I lead practical sessions that help teams design, evaluate, and refine AI systems, covering everything from intent mapping and feedback loops to prompting, evaluation, and ethical use.
UX Writing & InformationArchitecture
Shaping tone, structure, and language to make complex systems feel simple.
Talks & Keynotes
I believe the best way to meet the future is with understanding, not fear. My talks and keynotes help teams and audiences learn how to work with AI in ways that build confidence, curiosity, and care.
Recent work from across my AI systems practice.
At Zendesk, I helped redesign the moment between automation and live support to build trust in the system. Many users asked to speak to an agent too early, so we introduced gentle prompts that encouraged them to explore self-serve options first. These design shifts improved completion rates and model learning, and helped people feel more confident interacting with AI.
At Figma, I built a quality assurance program that helped teams understand and improve how AI performs. By connecting human feedback with model behavior, we were able to measure not just accuracy, but alignment and usefulness, creating a shared foundation for trust and improvement across teams.See Portfolio
Tools I use to structure problems, design AI behavior, and help teams work with more clarity.
I created a quality framework that turns abstract AI behavior into measurable signals. It helps teams evaluate how well systems understand, align, and respond, then connects those insights back to design and training decisions that make the AI better over time.
Explore the signals →
A practical model for tuning systems to respond with context and care. I use it to guide AI behavior reviews and prompt design: Alignment ensures intent is understood, Advocacy makes responses actionable, and Awareness calibrates tone and confidence.
View Framework →
A visual for the framework I use, that maps how conversation design changes across rule-based, hybrid, and LLM-driven systems—showing what stays constant (intent, context, user need) and what transforms (flow, data, feedback).
See the map →
Hi, I’m Marlinda
I help machines mind their manners.At Figma, I lead work in automation and AI design, creating frameworks that help technology learn from feedback and improve over time.Before that, I shaped content and AI systems at Zendesk, Adobe, Salesforce, and Yahoo. In every role, the throughline is the same: design is how intelligence learns to listen.When I’m not building, I share what I learn through talks, trainings, and essays about feedback, language, and the human side of AI.
CES 2025 – Emcee & Host, AI House
Leading sessions on AI trust, behavior, and human-centered systems.
Appeared in Forbes’ Women in AI coverage during CES 2025
See article
That’s how trust is built. That’s how intelligence is shaped. And that’s what I design.
I design AI systems by shaping how they reason, respond, and earn trust in practice.
Trusted by teams at:
I help teams design, build, and evolve AI systems that drive efficiency, consistency, and customer trust.
Conversation Design
Crafting natural, context-rich dialogues that help users feel understood and supported.
Consulting
I advise organizations on AI systems design, evaluation strategy, and feedback architecture. Whether you’re building from scratch or scaling an existing platform, I help connect intent, performance, and impact.
AI Evaluation Systems
I turn metrics into insight, spotlight optimization opportunities, and guide smarter model improvement.
Workshops & Trainings
I lead practical sessions that help teams design, evaluate, and refine AI systems, covering everything from intent mapping and feedback loops to prompting, evaluation, and ethical use.
UX Writing & InformationArchitecture
Shaping tone, structure, and language to make complex systems feel simple.
Talks & Keynotes
I believe the best way to meet the future is with understanding, not fear. My talks and keynotes help teams and audiences learn how to work with AI in ways that build confidence, curiosity, and care.
Recent work from across my AI systems practice.
At Zendesk, I helped redesign the moment between automation and live support to build trust in the system. Many users asked to speak to an agent too early, so we introduced gentle prompts that encouraged them to explore self-serve options first. These design shifts improved completion rates and model learning, and helped people feel more confident interacting with AI.
At Figma, I built a quality assurance program that helped teams understand and improve how AI performs. By connecting human feedback with model behavior, we were able to measure not just accuracy, but alignment and usefulness, creating a shared foundation for trust and improvement across teams.
Tools I use to structure problems, design AI behavior, and help teams work with more clarity.
I created a quality framework that turns abstract AI behavior into measurable signals. It helps teams evaluate how well systems understand, align, and respond, then connects those insights back to design and training decisions that make the AI better over time.
Explore the signals →
A practical model for tuning systems to respond with context and care. I use it to guide AI behavior reviews and prompt design: Alignment ensures intent is understood, Advocacy makes responses actionable, and Awareness calibrates tone and confidence.
View Framework →
A visual for the framework I use, that maps how conversation design changes across rule-based, hybrid, and LLM-driven systems—showing what stays constant (intent, context, user need) and what transforms (flow, data, feedback).
See the map →
Hi, I’m Marlinda
I help machines mind their manners.At Figma, I lead work in automation and AI design, creating frameworks that help technology learn from feedback and improve over time.Before that, I shaped content and AI systems at Zendesk, Adobe, Salesforce, and Yahoo. In every role, the throughline is the same: design is how intelligence learns to listen.When I’m not building, I share what I learn through talks, trainings, and essays about feedback, language, and the human side of AI.
CES 2025 – Emcee & Host, AI House
Leading sessions on AI trust, behavior, and human-centered systems.
Appeared in Forbes’ Women in AI coverage during CES 2025
See article
That’s how trust is built. That’s how intelligence is shaped.
And that’s what I design.
I design AI systems by shaping how they reason, respond, and earn trust in practice.
Trusted by teams at:
I help teams design, build, and evolve AI systems that drive efficiency, consistency, and customer trust.
Conversation Design
Crafting natural, context-rich dialogues that help users feel understood and supported.
Consulting
I advise organizations on AI systems design, evaluation strategy, and feedback architecture. Whether you’re building from scratch or scaling an existing platform, I help connect intent, performance, and impact.
AI Evaluation Systems
I turn metrics into insight, spotlight optimization opportunities, and guide smarter model improvement.
Workshops & Trainings
I lead practical sessions that help teams design, evaluate, and refine AI systems, covering everything from intent mapping and feedback loops to prompting, evaluation, and ethical use.
UX Writing & InformationArchitecture
Shaping tone, structure, and language to make complex systems feel simple.
Talks & Keynotes
I believe the best way to meet the future is with understanding, not fear. My talks and keynotes help teams and audiences learn how to work with AI in ways that build confidence, curiosity, and care.
Recent work from across my AI systems practice.
At Figma, I built a quality assurance program that helped teams understand and improve how AI performs. By connecting human feedback with model behavior, we were able to measure not just accuracy, but alignment and usefulness, creating a shared foundation for trust and improvement across teams.See Portfolio
At Zendesk, I helped redesign the moment between automation and live support to build trust in the system. Many users asked to speak to an agent too early, so we introduced gentle prompts that encouraged them to explore self-serve options first. These design shifts improved completion rates and model learning, and helped people feel more confident interacting with AI.
Tools I use to structure problems, design AI behavior, and help teams work with more clarity.
I created a quality framework that turns abstract AI behavior into measurable signals. It helps teams evaluate how well systems understand, align, and respond, then connects those insights back to design and training decisions that make the AI better over time.
Explore the signals →
A practical model for tuning systems to respond with context and care. I use it to guide AI behavior reviews and prompt design: Alignment ensures intent is understood, Advocacy makes responses actionable, and Awareness calibrates tone and confidence.
View Framework →
A visual for the framework I use, that maps how conversation design changes across rule-based, hybrid, and LLM-driven systems—showing what stays constant (intent, context, user need) and what transforms (flow, data, feedback).
See the map →
Hi, I’m Marlinda
I help machines mind their manners.At Figma, I lead work in automation and AI design, creating frameworks that help technology learn from feedback and improve over time.Before that, I shaped content and AI systems at Zendesk, Adobe, Salesforce, and Yahoo. In every role, the throughline is the same: design is how intelligence learns to listen.When I’m not building, I share what I learn through talks, trainings, and essays about feedback, language, and the human side of AI.
CES 2025 – Emcee & Host, AI House
Leading sessions on AI trust, behavior, and human-centered systems.
Appeared in Forbes’ Women in AI coverage during CES 2025
See article
