Custom AI Models
AI solves many of the questions we have in the world today. With Agriculture, many of the challenges in solving complex issues are rooted in data quality and data quantity. With TensoAI, we develop AI models to fit your needs, and work with you to achieve a successful outcome.
- Harvest Forecasting with Outdoor Crops
- Hops
- Wheat, Barley, Hemp (Q4 2022)
- Yield Prediction with Indoor Growing Spaces
- Various Lettuce cultivars
- Strawberries
- Peppers (Q2 2023)
- Quality Prediction for multiple crop types (in development)
- Hardware failure prediction – Anomaly detection on voltage readouts
The Seed Network
When most of the worlds plant data is held within data silos and data lakes, model development becomes challenging in creating solutions that scale. The Seed Network establishes the first Federated Learning Plant Ecosystem that develop AI solutions with privacy-preserving tools. Developing models with privacy as a priority unlocks many data silos and allows us to train models utilizing multiple growing techniques to prevent overfitting.
- Private secure network to train, test, and develop models on Plant Growth
- Both Tabular and plant vision available for specific crops
- Indoor Crops (Cannabis, Strawberries, Lettuce)
- Outdoor Crops (Barley, Wheat, Corn)
- Currently working with Tensorflow and Pytorch
Data Transformation
Unfortunately RAW data points become problematic and time consuming for ML engineers to manage. Data Pipeline (or MLOps) is a challenging field that is dependent on Engineers’ AI development. Save time, resources and money by developing with TensoAI.
- Working with all major databases
- Create Scripts for automatic data pipeline integration
- Work in AWS, Azure, local infrastructure and custom existing pipelines
- Satellite Imagery (MODIS, Sentinel-2)
- NDVI
- Biomass Index
- Soil Moisture (5 cm depth)
- Tabular Data Transformation for:
- Data Visualization and Analytics
- Model training, testing and development
How we do it
1. Advise and Blueprinting
Every growing environment is different. From different technologies being used to multiple growing methods utilized, it important to understand how data and AI will play a role in the end goal. TensoAI’s 3 step process identifies equipment, processes and data pipelines that describes client needs and general assessment of your Crop Business. A exploratory report is generated as a discussion tool to isolate specific plans in the next step.
2. Develop and Test
Planning is key to a successful project. Defining clear goals and milestones establishes key performance indicators in project management. TensoAI works with you in establishing the framework and delivering on the intended goals.
3. Continuous Testing and Scaling
Model Accuracy and Data Quality plays an important role for TensoAI. We create automated markers in key performance indicators and share with appropriate stakeholders for full clarity. TensoAI continuously tests and maintains its infrastructure and models for multiple challenges such as, over-fitting, data drift, and data drops.