Heads up: This is a quick showcase — a glimpse into my process, decisions, and outcomes.
Want the full story? Let’s talk.
How Agriculture Works
Jan-Feb
Planning
Digitally map fields, decide which crop(s) to plant.
Mar-Apr
Seeding
Decide seed type, quantity, placement, and timing.
May-Sep
Monitoring
Monitor crop growth and health throughout the season.
This Project
Oct-Dec
Analysis
Analyze data after crops are collected, plan for next season.
Context and Challenges
My Role: Lead Product Designer
Project: Field Monitoring Dashboard
Focus: Designed scalable insights to help reps monitor field health efficiently.
Users
Seed Sales Reps
Advice farmers and manage seed sales and decisions in the field.
Manages 50+ Farmers
30,000+ Acres per Rep
Painpoints
Must physically visit fields to find issues — not always feasible.
Can't visit every field in time
Cost and Impact
Late detection = lost crops and lost customers.
Up to 30% crop loss
Loss of customers
Strategy - Version 1
Leadership Goals
First use of the data science model in the field
What they wanted
How they planned to achieve it
Reduce the time reps spend physically checking fields
Use a data science model built over 5 years
Give them a prioritized list so they visit only the most critical ones
Generate a ranked list of fields based on urgency
Engineering Collaboration
Analyzed the model’s output, structure, and data frequency
Planned how insights should be presented clearly to reps
Key Decisions: Leadership & Engineering
The model was production-ready — but reps’ trust wasn’t validated.
Instead of overbuilding, we validated usability first.
Research and Key Insights
5 Reps Interviewed
Used early prototypes to test trust and actionability
Insights synthesized with Dovetail AI + ChatGPT
Reps found the priority list helpful — it helped them focus their field visits
But the model showed what’s wrong, not why
Reps wanted weather data to spot issues proactively
Solution Design - Version 1
Connecting Research to Decisions and Features
User Insight
How We Solved It
Reps struggled to prioritize field visits
Created a ranked field list, allowing reps to focus on the most critical fields first
Reps needed an at-a-glance view to assess which fields needed urgent attention
Built a risk-based alert system to flag critical fields and help reps take faster action
Reps needed field health indicators to understand why a field wasn’t performing well
Added field-level indicators to help diagnose potential causes of poor performance
Trade-offs and Constraints
Version Key: V1 = Initial Launch, V2 = Future Enhancements
Features
V1
V2
Reasoning
Ranked Field List |
Core functionality — reps needed direction to avoid visiting fields blindly
|
Critical user need — helped reps act faster by flagging urgent fields |
Field Health Indicators
|
High-priority request — helps assess field conditions and growth |
|
|
|
|
|
|
Visual Evolution - Version 1
Visual Evolution - Version 1
Main UI Screen
The dashboard highlights priority fields, alerts, and field health indicators — all in one actionable view at scale.
Risk based alert.
Key health indicator - Growing Degree Units
Ranked Field List
Key Iterations
Before
Main UI
Map-based layout, visually striking but resembled a discontinued tool — raised perception concerns among users
After
Decision Criteria - Main UI
Switched to a clean grid layout aligned with existing tools — improved trust and usability
Before
Cards
Health data was displayed in long-form cards — too text-heavy and hard to scan.
After
Decision Criteria - Cards
Redesigned into compact data blocks — supports quick comparison and decision-making
Measurable Results - Version 1
15% adoption
within first 2 months.
Strong early uptake — even with limited promotion
8 of 10 reps
expressed interest in continued use
“I’d pull this up before a sales call to see what’s going on.” — User Quote
3 new regions
planned expansion by leadership
Early success led to prioritization for scale across additional U.S. territories
What Worked
Reps appreciated the simplicity of the ranked list and alerting system — it gave them a fast, focused way to spot fields that needed attention.
What Was Missing
Frequently used indicators were still missing, even though V1 covered the some critical ones
More visual ways to show both the overall list and individual indicators were needed, so reps could absorb information at a glance.
More ways to organize or prioritize fields were expected by reps — even if the layout remained the same.
Why We Iterated
In Version 2, we shifted from simply saying “this field needs attention” to helping reps understand
why it matters — and what to do next.
Strategy - Version 2
Leadership Direction
Leadership wanted to expand the tool by adding more field-level indicators, based on strong early adoption
They also proposed integrating features from other internal models to increase overall platform value
Engineering Collaboration
Engineering built a dedicated weather data pipeline, enabling deeper insights and easier integration of future indicators
We collaborated with the design systems team to explore scalable and visually intuitive presentation methods
Key Decisions: Leadership & Engineering
We shifted the focus in Version 2 from adding more data to creating clearer context around field issues
To support this shift, we prioritized a second round of user research before building new functionality
Research and Key Insights
8 Reps Interviewed
Evaluated how well V1 indicators worked in real-world decision-making
Explored which indicators reps trusted, used, or ignored
Synthesis done using Dovetail AI + ChatGPT
Reps needed to see multiple indicators together (e.g. weather + disease), not in isolation
They wanted to visually track how field conditions changed over time
They needed a more visual, at-scale view of both field conditions and seed performance — all in one place
Solution Design - Version 2
Connecting Research to Decisions and Features
User Insight
How We Solved It
Reps needed a broader view beyond weather
Added more health indicators for a fuller picture |
Reps wanted clearer trend visualization
Introduced graph and map-based views for faster scanning |
Reps wanted to access existing insights trapped in other internal tools
Integrated those features to make the tool self-contained |
Reps wanted to understand how seed products influenced field performance
Linked outcomes to seed product use for better context
Trade-offs and Constraints (V2 -> V3)
Features
V2
V3
Reasoning
|
Helped reps recognize spatial patterns and regional stress clusters |
|
|
|
|
|
|
|
|
|
|
|
|
Field Health Indicators
|
|
|
|
|
|
|
|
|
|
|
|
Visual Evolution - Version 2
Main UI Screen
The new dashboard brings together weather graphs, map views, and seed product view — making field insights more visual, contextual, and actionable in one place.
Weather Graph
Map Views + Disease Pressure
Notes
Seed Product Views
Key Iterations
Before
Map View
Map and table sat side by side to show data and location, but lacked clarity and scale control.
After
Decision Criteria - Map View
Redesigned to support zooming, scaling, and easier data readability — reflect past preferences from the commercial team.
Before
Mobile UI
Mobile used a mirrored layout of the PWA, making navigation slow and cluttered.
After
Decision Criteria - Mobile UI
Mobile experience was rebuilt for on-the-go use, with simplified layouts and field-first prioritization.
Before
Notes
Notes were displayed as a summary, similar to other field indicators.
After
Decision Criteria - Notes
Notes included a lot of content and images, so a separate modal was introduced to keep everything clear and focused.
Measurable Results - Version 2
38% adoption
up from 15% in Version 1
The shift to mobile access, deeper insights, and visual improvements drove stronger engagement in V2.
64% used mobile daily
accessing the tool at least once per day
“First thing in the morning, I open it on my phone to check the weather before heading out." - User Quote
70% used graph views
during their first visit
The new GDU and precipitation visuals made it easier to interpret field trends right away.
Key Learnings
Designing this tool taught me that clarity beats complexity — especially for reps in the field. A model is only useful if people trust it, so we focused early on building credibility, not just shipping features.
I also learned to design for scale without overcomplicating. Every choice in V2 balanced future flexibility with immediate usability — clean layouts, clear signals, minimal effort.
Personal Growth
This was a shift from just designing screens to shaping how data becomes action. I worked across product, engineering, and science teams — translating raw metrics into real-world value.
It pushed me to think like a systems designer: scalable layouts, mobile-first decisions, and patterns built to evolve.
What's Next
We’re now building Version 3, launching in early 2026.
It’ll connect seed choices to performance, let reps track changes over time, and surface insights they can act on — before issues even start.
We’re also exploring ways to bring AI into the workflow — not as a chatbot, but to highlight patterns, flag anomalies, and help reps focus on what matters most.
Version 1 proved value.
Version 2 gave context.
Version 3 is all about driving decisions — smarter and faster.
