Big Data Annotation for Dutch AgriTech Company
Ongoing image annotation of fruits and vegetables
Netherlands
Model of cooperation:
2 Full-time Data Annotators and 2 Part-time Data Annotators.
Ongoing annotation of
thousands
of images of fruits and vegetables monthly
Cooperation Background
An AgriTech company from the Netherlands specializing in computer vision systems develops advanced solutions for automated inspection and grading of fruits and vegetables. Their technology evaluates produce based on size, shape, color, and defects, enabling producers to streamline quality control processes.
As part of improving their machine learning models, the company required high-quality image annotation to support complex object detection tasks. To ensure scalability and maintain focus on core development, they opted for business process outsourcing.
Client’s Problem
The client required a dedicated annotation team to work on image datasets using polygon-based labeling. The project involved annotating fruits and vegetables with high precision, which demanded both technical expertise and domain understanding.
In addition, the client had several specific requirements:
- Annotators needed to work within the client’s custom annotation tool
- Direct communication between annotators and the client was necessary for efficiency
- Consistency in the team was critical to reduce onboarding time and maintain quality
Project Challenge
The annotation tasks were complex and required attention to detail, as polygon-based labeling involves accurately outlining irregular shapes. This made it essential to select specialists with prior experience in similar projects.
Moreover, integrating annotators into the client’s internal workflow and tools required careful onboarding. Maintaining consistent communication while ensuring productivity and quality added another layer of complexity.
Mobilunity-BPO Solution
Mobilunity-BPO organized a tailored selection process to identify candidates with relevant experience in image annotation, particularly in projects involving agricultural products and polygon labeling.
The process included
- Conducting personal interviews with shortlisted annotators together with the client
- Assigning test tasks to validate skills and accuracy
- Selecting a team of two full-time and two part-time annotators
Once selected, the team was:
- Fully onboarded into the client’s internal annotation system
- Trained to align with project-specific requirements and workflows
To ensure transparency and smooth collaboration, Mobilunity-BPO also provided:
- Weekly reports detailing hours worked and types of completed tasks
- Ongoing coordination and support throughout the project
Outcome
The client received a dedicated and consistent annotation team that met all project requirements from the start. Thanks to careful selection and onboarding, the team quickly adapted to the client’s processes and tools.
Overall, the client was able to enhance their computer vision models without diverting internal resources, ensuring faster development cycles and reliable results.
The collaboration resulted in:
- High-quality, precise annotations aligned with model requirements
- Improved efficiency through direct communication between annotators and the client
- Reduced onboarding time due to team consistency
Few Words About Cooperation with Us
Our priority was to build a consistent, skilled annotation team that could meet strict quality requirements from day one. This approach allowed the client to accelerate model development without compromising on precision
