Product Design

App

AI

LORE

AI assistant for farmers

Product Design

App

AI

LORE

AI assistant for farmers

Product Design

App

AI

LORE

AI assistant for farmers

Prodap, now dsm-firmenich, is a company in the agricultural sector that offers, among its products, digital solutions for large beef cattle farms in Brazil.

Company

PRODAP

Year

2021

Product

LORE

Role

Product Designer

Duration

One year

Task

Research, prototype and deliver a MVP.

Tools

User Interviews, Design Sprint, Personas, Value Matrix, User Flows, Figma

Target Users

Managers and owners of cattle farms.

Results

The first AI assistant for farmers and SEK 70 million invested in the app.

Problem


One of the biggest challenges faced by managers and owners of these large cattle farms is keeping daily track of operations issuces and production results due to the vast amount of data generated every day. They have many other tasks in their daily routines and need quick, straightforward access to this information - focused only on what they truly need and care about.

Problem


One of the biggest challenges faced by managers and owners of these large cattle farms is keeping daily track of operations issuces and production results due to the vast amount of data generated every day. They have many other tasks in their daily routines and need quick, straightforward access to this information - focused only on what they truly need and care about.

Responsibilities

During the Design Sprint process, I collaborated with two other UX designers from the company. As for the product itself, I was responsible for its entire interface, from the conversation to the style guide, graphics, information architecture, icons and tables. I was also responsible for interviewing users and presenting the feedback to the company, and defining the next steps of the project.

Research Process


The first step in this investigation process was a Design Sprint. In 5 days, cattle consultants and farm directors together the PRODAP team of designers, project managers and costumer success joined this ideation phase.

Research Process


The first step in this investigation process was a Design Sprint. In 5 days, cattle consultants and farm directors together the PRODAP team of designers, project managers and costumer success joined this ideation phase.

During the ideation phase, the group defined the problem to be addressed, identified the stakeholders, established the criteria for the product's success, and outlined the potential benefits for users. This is how the first ideas emerged.

Then, two key personas were created: a farm director and a people's manager. One is only interested in financial results, while the other wants an overall overview of the activities.

Then, two key personas were created: a farm director and a people's manager. One is only interested in financial results, while the other wants an overall overview of the activities.

Prioritization


This discovery was important for prioritizing the product's features. For this, the Value vs. Effort Matrix was used, where ideas were mapped based on the Design Sprint team's assessment. The ideas that would require less effort to implement while delivering the most value to users were prioritized.

Prioritization


This discovery was important for prioritizing the product's features. For this, the Value vs. Effort Matrix was used, where ideas were mapped based on the Design Sprint team's assessment. The ideas that would require less effort to implement while delivering the most value to users were prioritized.

These where the priorities and opportunities:


Top user priorities


  • Reports and graphs on animal feed consumption

  • Performance results from each farm area

  • Overall analysis of farm operations


Additional features


  • Quick messages (e.g., alerts, reminders)

  • Practical tips for farm management


Strategic opportunity for the company


  • Leverage artificial intelligence (AI) to address these needs

  • Integrate AI into clients’ digital transformation process

These where the priorities and opportunities:


Top user priorities


  • Reports and graphs on animal feed consumption

  • Performance results from each farm area

  • Overall analysis of farm operations


Additional features


  • Quick messages (e.g., alerts, reminders)

  • Practical tips for farm management


Strategic opportunity for the company


  • Leverage artificial intelligence (AI) to address these needs

  • Integrate AI into clients’ digital transformation process

Solution

Our solution is a AI-powered tool to automate and enhance what farm managers and owners need: data visualization and actionable insights.

Business Alignment

As part of the design strategy for LORE, we identified an opportunity to integrate data from Prodap Views Master, an existing monitoring system used by many of our clients. By using its data infrastructure to power LORE’s dashboards and insights, we achieved two key business outcomes:

  • Increased product synergy: LORE became a natural extension of Views Master, encouraging clients to adopt both solutions.

  • Accelerated sales cycle: The integration added immediate value for existing customers, reducing onboarding time and making the offering more attractive.

From a UX perspective, this also improved consistency across platforms. I worked closely with product managers and data engineers to ensure the design supported both user needs and business goals.

Business Alignement

As part of the design strategy for LORE, we identified an opportunity to integrate data from Prodap Views Master, an existing monitoring system used by many of our clients. By using its data infrastructure to power LORE’s dashboards and insights, we achieved two key business outcomes:

  • Increased product synergy: LORE became a natural extension of Views Master, encouraging clients to adopt both solutions.

  • Accelerated sales cycle: The integration added immediate value for existing customers, reducing onboarding time and making the offering more attractive.

From a UX perspective, this also improved consistency across platforms. I worked closely with product managers and data engineers to ensure the design supported both user needs and business goals.

MVP

LORE, the first AI-powered virtual assistant for livestock management, provides a digital platform for monitoring, diagnosing, and predicting farm activities. Users can interact with LORE via text or voice, accessing data in charts and tables for easy analysis.

MVP

LORE, the first AI-powered virtual assistant for livestock management, provides a digital platform for monitoring, diagnosing, and predicting farm activities. Users can interact with LORE via text or voice, accessing data in charts and tables for easy analysis.

MVP

LORE, the first AI-powered virtual assistant for livestock management, provides a digital platform for monitoring, diagnosing, and predicting farm activities. Users can interact with LORE via text or voice, accessing data in charts and tables for easy analysis.

Implement AI in LORE was a decision to enhance efficiency in livestock management by automating data analysis and decision-making. By leveraging AI, LORE processes large volumes of real-time data, identifies patterns, and provides precise recommendations, allowing farmers to optimize resources and improve overall productivity.

Implement AI in LORE was a decision to enhance efficiency in livestock management by automating data analysis and decision-making. By leveraging AI, LORE processes large volumes of real-time data, identifies patterns, and provides precise recommendations, allowing farmers to optimize resources and improve overall productivity.

Pilot Clients

For MVP testing, pilot customers were selected. Following product testing, we conducted two feedback interview rounds: one at the 3-month mark and another after 6 months of usage.

For the 3-month follow-up research, we invited users from nine different client organizations to participate in ~30 minute calls. The interviews followed a semi-structured format, covering:

  • Their daily routines

  • Current methods for collecting and recording farm data

  • Phone usage habits (e.g., notification preferences, internet connectivity)

  • Frequency of LORE app access

For the 3-month follow-up research, we invited users from nine different client organizations to participate in ~30 minute calls. The interviews followed a semi-structured format, covering:

  • Their daily routines

  • Current methods for collecting and recording farm data

  • Phone usage habits (e.g., notification preferences, internet connectivity)

  • Frequency of LORE app access

The objectives of the interviews were:

  • To align customer expectations regarding the product, comparing them with the value we're already delivering;

  • To gain deeper insights into users' daily routines and workflows;

  • To identify pain points, brainstorm solutions, and suggest potential improvements that could benefit them;

  • To validate whether we're on the right track.

The findings and feedback gathered from all 10 interviews were delivered to the company, as demonstrated in the following slides:

Considering these attention points, next steps were defined:

For the 6-month follow-up research, we invited users from sixteen different client organizations to participate in ~30 minute calls. The interviews again followed a semi-structured format, but this time we divided the clients into two groups: frequent LORE users and those who initially used LORE but later discontinued access.

To monitor individual client usage frequency, we analyzed data in Tableau:

For the 6-month follow-up research, we invited users from sixteen different client organizations to participate in ~30 minute calls. The interviews again followed a semi-structured format, but this time we divided the clients into two groups: frequent LORE users and those who initially used LORE but later discontinued access.

To monitor individual client usage frequency, we analyzed data in Tableau:

After analyzing the interviews and their results, the following next steps were defined:

After analyzing the interviews and their results, the following next steps were defined:

Lessons Learned

  • Writing clear, concise messages for an AI assistant is harder than it seems—especially for users in rural or low-tech contexts. After receiving feedback that 70% of users found the messages too long or confusing, I enrolled in a UX Writing course to sharpen my content design skills.

  • The interviews with pilot clients were crucial. The feedback we gathered directly shaped the assistant’s tone and flow.

  • Working with farm specialists taught me how to ask better questions and adapt design language to different knowledge levels.

Lessons Learned

  • Writing clear, concise messages for an AI assistant is harder than it seems—especially for users in rural or low-tech contexts. After receiving feedback that 70% of users found the messages too long or confusing, I enrolled in a UX Writing course to sharpen my content design skills.

  • The interviews with pilot clients were crucial. The feedback we gathered directly shaped the assistant’s tone and flow.

  • Working with farm specialists taught me how to ask better questions and adapt design language to different knowledge levels.

Lessons Learned

  • Writing clear, concise messages for an AI assistant is harder than it seems—especially for users in rural or low-tech contexts. After receiving feedback that 70% of users found the messages too long or confusing, I enrolled in a UX Writing course to sharpen my content design skills.

  • The interviews with pilot clients were crucial. The feedback we gathered directly shaped the assistant’s tone and flow.

  • Working with farm specialists taught me how to ask better questions and adapt design language to different knowledge levels.

Impacts

Following the pilot testing, LORE was added to dsm-firmenich's portfolio of digital products, proving to be a success story - as demonstrated by this client testimonial in the video below (in Portuguese):

"It's a fantastic application that shows you exactly what is happening on the property, even when you're not there. Decisions are made quickly because I have the data in the palm of my hand to plan the best strategy and achieve the planned objectives," says Ricardo Cezar Espírito Santo, farm Bom Sucesso owner.

Recognizing the potential of LORE and the implementation of artificial intelligence in livestock farming, Royal DSM acquired PRODAP, now called dsm-firmenich. Additionally, 70 million Swedish kronor were invested in the app over a two-year period.

Recognizing the potential of LORE and the implementation of artificial intelligence in livestock farming, Royal DSM acquired PRODAP, now called dsm-firmenich. Additionally, 70 million Swedish kronor were invested in the app over a two-year period.

View More Projects