Quick facts
| Field | Value |
|---|---|
| Role | Ideation, architecture, and development |
| Delivery | Web platform + AI |
| Stack | React Native, REST API, AI |
| Outcome | 1st place phase 1 · 3rd place phase 2 |
| Period | 2025 |
/ CONTEXT
Campo Inteligente was born during the RESTICII software residency in Bahia, as part of a challenge to apply AI to decision automation. After researching real-world problems, agriculture stood out: small farmers often decide what to plant with little structured information. I later developed the concept with an agronomist and validated it with people who work in the field.
/ CHALLENGE
Small farmers make planting decisions with limited data, leading to unsuitable crops, low predictability, and difficulty comparing scenarios. The challenge was to make AI-driven agricultural planning both scalable and accessible—not just for large farms, but for the people who need it most.
/ PROCESS
After market research, I defined the MVP with the team and designed a REST API that serves data and recommendations, an AI model inside the API, and a React Native mobile front end. The priority was a demonstrable product with clear flows and an easy-to-communicate value proposition.
/ SOLUTION
- Property and field records with soil and location data.
- AI recommendations for crops, planting windows, and crop combinations.
- Scenario simulations comparing yield and risk.
- Continuous climate alerts to support field decisions.
/ HIGHLIGHTS
1st place in the hackathon’s first phase
3rd place in the second phase
Top 3 at Bahia Innovate Summit 2025
/ RESULTS
The MVP stood out for combining social impact, AI, and a clear value proposition. It validated the concept with judges and strengthened its potential to support family farming, while consolidating my experience in product ideation, AI architecture, and delivery under pressure.
/ TECHNICAL
The solution combines a REST API, a React Native mobile front end, and an AI model hosted within the API. Keeping model training and inference in the API simplified the MVP architecture and maintained one source of truth for data and decisions.