Nelson Pinheiro

Agri-tech · ML systems · Remote sensing

Buildingintelligentsystemsforagriculture.

I combine deep domain knowledge in crop science with modern software engineering — building ML pipelines, geospatial platforms, and production SaaS products that solve real problems in agriculture.

Work with meView projects
// What I Do

What I Do

Three interconnected areas of expertise that I bring to every project.

4 projects

ML & Computer Vision

Deep learning pipelines for plant phenotyping, image segmentation, and classification systems.

View projects
3 projects

Modeling & Simulation

Mathematical models and simulations for soil-plant systems, rhizosphere dynamics, and growth prediction.

View projects
4 projects

Research Engineering

Reproducible pipelines, CI/CD for research, containerization, and deployment of scientific tools.

View projects
// Featured Work

Featured Work

Selected projects showcasing my approach to solving complex problems.

Live at nrate.farm — Germany's first economics-driven N calculator
MR=MC optimization model across 6 DüV-compliant crops
Bilingual (DE/EN) with locale-aware content segmentation
Research EngineeringLive

NRate.farm — Smart Nitrogen Calculator for Farmers

Full-stack agri-tech SaaS for economics-driven nitrogen fertilization decisions. Germany's first N-rate optimizer that balances marginal grain revenue against fertilizer cost — DüV-compliant, bilingual (DE/EN). Deployed at nrate.farm with active users.

agri-technextjstypescripti18n+3
Next.js 16TypeScriptTailwind CSS v4Framer Motion
Complete IoT monitoring solution
Prometheus/Grafana integration
Real-time VPD alerting
Research EngineeringCompleted✨ Interactive Demo

GrowController: IoT Environmental Monitor

ESP32-based CO2, temperature, and humidity monitor with VPD calculation and Prometheus/Grafana integration.

pipelinedeploymentcipython+1
C++ArduinoESP32Prometheus

Research in progress

Research EngineeringEarly Stage

EUDR Compliance Verification Engine

Satellite-based deforestation risk verification platform for EU Deforestation Regulation compliance. Combines Sentinel-2 imagery analysis with supply chain traceability ahead of the December 2026 enforcement deadline.

eudrremote-sensingcompliancesentinel-2
Next.jsPythonSentinel-2Details coming soon

Research in progress

ML & Computer VisionEarly Stage

Geospatial Risk Intelligence for Agricultural Lenders

Satellite-derived field risk scoring platform for agricultural credit assessment. Provides lenders with crop health indices, drought exposure, and yield stability signals at parcel level.

geospatialrisk-scoringremote-sensingfintech
Next.jsPythonSentinel-2Details coming soon
Unlimited perfectly-labeled training data (zero annotation cost)
1.536M-ray LiDAR point clouds for 3D structure
Pixel-perfect multispectral segmentation masks (RGB/NIR/Red-Edge)
Modeling & SimulationCompleted

GPU-Accelerated Synthetic Data Generation for Crop Phenotyping

Physics-based sensor simulation using C++ ray-tracing to generate perfectly labeled training data. Dual virtual sensors (LiDAR + multispectral camera) provide structural point clouds and pixel-perfect segmentation masks for bean-wheat intercropping ML research.

ray-tracingsensor-simulationsynthetic-datagpu+2
C++PythonHeliosCUDA
Quantified mobility-persistence trade-off across 4 soil types
Sand: HIGH leaching risk (low Kf, rapid peak)
Humus-rich: HIGH persistence (strong sorption, slow release)
Modeling & SimulationCompleted

Mechanistic Modeling of AMPA Persistence Across Soil Types

Meta-analysis using TESFO mechanistic modeling to evaluate glyphosate metabolite (AMPA) behavior across four soil types. Identified critical mobility-persistence trade-off requiring site-specific agricultural risk management.

mechanistic-modelingenvironmental-sciencesoil-physicsrisk-assessment+1
PythonTESFO ModelFreundlich IsothermLangmuir Isotherm
98.1% nitrogen accuracy (88.1% → 98.1% via PCA)
R²=0.844 water prediction, MAE=1.55%
1.8 MB edge-ready: 1440× faster than lab methods
ML & Computer VisionCompleted

Rapid Soil Water & Nitrogen Prediction via NIR Spectroscopy and ML

Non-invasive dual-model ML pipeline delivering soil predictions in <1 minute vs. traditional 24-48 hour lab assays. Combines SVR for water content (R²=0.844, MAE=1.55%) and Random Forest for nitrogen classification (98.1% balanced accuracy). Breakthrough: PCA reduced features 90% while increasing accuracy from 88.1% to 98.1%. Production-ready with OOD detection and confidence thresholding. Edge-optimized at 1.8 MB.

svrrandom-forestpcanir-spectroscopy+3
Pythonscikit-learnSVRRandom Forest
79,125 farms analyzed with dual spatial validation
Statistically significant clustering confirmed (p<0.05)
IDW and Kriging interpolation implemented
Modeling & SimulationCompleted

Nitrogen Loss Risk Spatial Analysis

Nationwide spatial risk mapping of German organic farms using Local Moran's I and Getis-Ord Gi* hotspot analysis. Cleaned 79,125 farm records, implemented IDW and Ordinary Kriging interpolation, and proved statistically significant clustering (p<0.05) linking sandy soils to high-risk zones.

geospatialspatial-statisticsgispolicy+2
PythonPySALGeoPandasQGIS
0.96 F1-score with PCA + Random Forest
63,650 spatially-indexed patches (geographic zone splits)
Solved overfitting: OOB-Test gap reduced from 0.12 → 0.03
ML & Computer VisionCompleted

Multispectral UAV Crop Classification Pipeline

End-to-end ML pipeline for temporal crop classification using 10-band UAV orthophotos. Built rigorous spatial zone-based data splits to prevent leakage, engineered 220 spectral features, and solved overfitting through PCA dimensionality reduction—achieving 0.96 F1-score with just 9 components.

classificationremote-sensinguavpca+3
Pythonscikit-learnRandom ForestPCA
View All Projects
// How I Work

How I Work

01

Scientific Rigor

Every model is grounded in domain knowledge and validated against real data.

02

Reproducibility

Version-controlled code, containerized environments, and documented workflows.

03

Clean Engineering

Type-safe code, tested pipelines, and maintainable architectures.

04

Collaboration

Clear documentation, modular design, and open communication.

Let’s Work Together

Available for select consulting engagements. I bring both domain knowledge and engineering depth to agri-tech products.

Work with meContact Me