Hydra Efficacy Assessment Program (H.E.A.P)

The Hydra Efficacy Assessment Program (H.E.A.P) is a first of its kind initiative directed towards food organizations to enable them to build the foundation towards realizing the dream of an automated food quality assessment process.

The HEAP Timeline

Phase I: AI Validation

Data collection and AI development

The partner organization will deploy food inspectors to collect and annotate data for the defects agreed upon.

Measure accuracies

AI efficacy can be proven if the initial accuracies are above 60%. This means that the AI models can be optimized to improve the accuracy.

Phase II: AI Optimization


The second phase involves optimizing the AI models. At the end of Phase II, AgShift will provide results on whether Hydra can be used to evaluate defects of the selected commodity and if the solution is reproducible, objective, and rapidly scalable.

“The defect variation across commodities is endless. An AI solution designed to work with strawberries does not work with raspberries. HEAP allows companies to understand how AI is a viable alternative to analyze and grade the commodities they work with, but the ones we have not addressed yet.” – Miku Jha, Founder & CEO, AgShift

Why become a partner?

AI viability assessment

Understand whether automation is the way forward in your food inspection process.

Low financial risk

Test the waters without making any major investment in new technologies.

Long-term benefits

HEAP partners receive  25% off on any commercial products or services procured by them from AgShift for three years.

How does it work?

1. Know Your Commodity

The focus in this step will be to understand what defects exist in your commodity, the criticality of each defect and their occurrences.

2. Engineering Assessment

Our scientists will attempt to understand if the defects are identifiable consistently using machine learning algorithms.

3. Partnership Agreement

Upon signing of an agreement, AgShift will officially induct your commodity into its Commodity of Interest list and kick-off HEAP implementation.

4. Hydra Deployment

A Hydra analyzer will be deployed at the partner’s facility to enable data collection.

5. Data Collection and Analysis

The inspectors will start collecting data and the AI scientists will use them to build models.

6. Evaluation and Acceptance

The efficiency of AI solution developed over the partnership period is shared with the partner.