Mobile interface for an agentic AI gardening companion, designed to keep the agent’s data fresh.

Four screens from the card-collection flow: a check-in card on Home, one question, cards to save, and saved cards two days later.
Four screens from the card-collection flow: a check-in card on Home, one question, cards to save, and saved cards two days later.

OVERVIEW

Strengthening the feedback loop between the gardener and the agent.

Growzi is an agentic AI gardening coach for iOS. The agent’s predictive value scales with user data, but it relies entirely on active input. As a product designer on a cross-functional team, I designed the mediation layer, collapsing the friction of data entry into lightweight, routine interactions.

IMPACT

Designed lightweight logging interactions that cut user friction and kept plant data accurate.

TEAM

1 stakeholder, 1 product manager, 3 designers, 4 data scientists

YEAR

2026

IMPACT

Designed lightweight logging interactions that cut user friction and kept plant data accurate.

TEAM

1 stakeholder, 1 product manager, 3 designers, 4 data scientists

YEAR

2026

YEAR

2026

THE PROBLEM

An empty agent cannot give valuable advice.

Unlogged daily events cause the digital record to fall behind the physical garden, making the agent's advice obsolete. The core challenge is data capture rather than processing, as only the gardener knows exact harvests or subtle changes.

Illustration: Growzi's garden record stays flat while the real garden drops with each unlogged harvest, opening a gap between them.

THE CHALLENGE

Designing for dirty hands and clean data.

An empty agent cannot reason effectively. If everyday events go unlogged, the system fails. The design challenge lay in creating a compelling, invitation-first relationship where brief micro-interactions effortlessly close the gap between the real garden and the digital record.

THE RELATIONSHIPS

Designing for dirty hands and clean data.

An empty agent cannot reason effectively. If everyday events go unlogged, the system fails. The design challenge lay in creating a compelling, invitation-first relationship where brief micro-interactions effortlessly close the gap between the real garden and the digital record.

Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.
Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.

What the gardener sees

Everything, in the moment. None of it written down.

Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.
Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.

Growzi's record

The agent reasons over this record and outside data like weather. It can estimate, but it can’t observe.

THE LOOP

Tightening the loop through progressive inputs.

The system moves through four phases: establishing context via home screen check-ins, resolving ambiguity through processed natural language, predicting outcomes with updated forecasts, and guiding action when insights are resurfaced.

01.

Establish context

Growzi’s agent notices nothing’s been logged during a busy stretch and asks for a quick look.

02.

Resolve ambiguity

The gardener answers in one sentence. The agent plays back what it heard (“About 8–10 lbs you didn’t log”) so nothing is guessed.

03.

Predict outcomes

The agent updates the forecast: 4–5 days of prime fruit left. The gardener saves it for later.

04.

Guide action

Two days later, the saved card is still current, and “Plan my gives” turns it into action.

PRINCIPLE 01

Log in passing.

Growzi’s agent capitalizes on fragmented inputs. The home screen widget lowers the data-entry threshold to a single tap or a raw verbal sentence, capturing context immediately without requiring an open app session.

Three steps: a lock-screen widget asks "Picked anything today?", the gardener sends a one-sentence voice note, and it is saved to the fruit tree's harvest log.
Three steps: a lock-screen widget asks "Picked anything today?", the gardener sends a one-sentence voice note, and it is saved to the fruit tree's harvest log.

PRINCIPLE 02

Ask when it matters.

Gardeners won't remember to log, and the agent can't wait forever. When the agent notices nothing's been logged during a busy stretch, it asks through a check-in card on Home. It fills in what it can, asks one question at a time for what's missing, and nothing is logged until the gardener confirms.

Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.
Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.

Isolate attention

Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.
Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.

Protect focus

Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.
Three annotated screens: a check-in card on Home, one question answered with a tap, and the agent playing back its rough estimate before anything is logged.

Guarantee data integrity

PRINCIPLE 03

Make responding feel low-stakes.

User participation remains optional through asynchronous, dismissible cards that adapt to physical workflows without demanding rigid compliance.

TAKEAWAYS

What I learned about designing for AI-native products.

01

Designing for AI means designing a relationship

An agent is only as good as what people tell it. The interface is the deciding factor in whether sharing that data feels like a chore or a natural interaction.

02

The stakes are higher for data integrity.

Every interaction feeds the model the agent learns from. A confusing screen doesn't just frustrate a user, it can lead to bad inputs.

03

Traditional UX fundamentals still apply.

Systems still need to anchor to predictable human mental models rather than forcing adaptation.

TAKEAWAYS

What I learned about designing for AI-native products.

01

Designing for AI means designing a relationship

An agent is only as good as what people tell it. The interface is the deciding factor in whether sharing that data feels like a chore or a natural interaction.

02

The stakes are higher for data integrity.

Every interaction feeds the model the agent learns from. A confusing screen doesn't just frustrate a user, it can lead to bad inputs.

03

Traditional UX fundamentals still apply.

Systems still need to anchor to predictable human mental models rather than forcing adaptation.

TAKEAWAYS

What I learned about designing for AI-native products.

01

Designing for AI means designing a relationship

An agent is only as good as what people tell it. The interface is the deciding factor in whether sharing that data feels like a chore or a natural interaction.

02

The stakes are higher for data integrity.

Every interaction feeds the model the agent learns from. A confusing screen doesn't just frustrate a user, it can lead to bad inputs.

03

Traditional UX fundamentals still apply.

Systems still need to anchor to predictable human mental models rather than forcing adaptation.

REFLECTION

How our early-stage team explored with AI.

These are just a few of the ways we approached the challenge. An AI-native app brings a new range of problems, and the work shifts from designing interfaces people navigate to designing systems people navigate with an agent. The team was early-stage and scrappy, and the process was mostly trial and error. AI-powered workflows let us iterate and explore many directions quickly.

Back to selected work