Junior Data Scientist
FieldPro · Kalama ward, Machakos County, Kenya
Oil and Gas · 2-10 employees
About the role
Develop, test, and refine computer vision models for SKU-level shelf recognition and merchandising KPIs. Collaborate with cross-functional teams to deploy models into production and manage client-ready reporting.
What they look for
Requirements
Requires strong proficiency in Python, SQL, and deep learning libraries such as PyTorch and YOLO. Candidates should have initial experience with object detection or image recognition projects.
Benefits
Full description
About Optimetriks - FieldPro
We are an advanced SaaS platform leveraging AI and automation to optimize field operations. We empower businesses with real-time data collection, predictive analytics, and smart workflow automation, enabling seamless management of field teams and operations.
Designed for industries like FMCG, financial services, and Agricultural sector, we transform field execution into a data-driven, high-performance engine.
Website: https://fieldproapp.com/
Requirements
We’re seeking a junior Machine Learning Engineer to power FieldPro Detect, our image recognition solution for merchandising operations. Your role: build and improve the computer vision models that turn store shelf photos into actionable insights — SKU detection, share of shelf, on-shelf availability, planogram compliance, price reading — for English- and French-speaking clients. As part of our AI team, you’ll collaborate with Data Scientists, Developers, Labelling, and Operations to drive impactful projects.
About the role
- Develop, test, and refine object detection and image classification models (YOLO family) for SKU-level shelf recognition.
- Train and fine-tune models on client SKU catalogues, tracking datasets and experiments in MLFlow.
- Improve model robustness to real field conditions: lighting, occlusions, dense shelves, low-end cameras.
- Accelerate annotation with ML-assisted pre-labelling (YOLO, SAM2) and collaborate with the Labelling team on quality in Label Studio.
- Build and maintain versioned training datasets and automated data pipelines.
- Deploy and serve models on our inference stack (NVIDIA Triton, Flask API) and ensure smooth releases to production.
- Maintain model monitoring and performance tracking dashboards.
- Turn detections into merchandising KPIs: share of shelf, on-shelf availability, planogram compliance, price reading.
- Take part in client onboardings: SKU catalogue scoping, labelling batches, training, validation, and go-live.
- Prepare client-ready reports and presentations explaining model outputs and insights.
About you
- Strong taste for deep learning and computer vision
- 100% proficient in Python and SQL
- Knowledge of ML/DL libraries like PyTorch, Ultralytics/YOLO, OpenCV, Scikit-learn
- A first experience with object detection or image recognition projects
- Team player, self-motivated, energetic
- Proficient in English (French is a plus)
What we offer
- Position: Junior Machine Learning Engineer – Computer Vision
- Contract: full-time, permanent (CDI)
- Salary: based on experience
- Real responsibilities with client interactions and operational deployments
- High growth environment
- Young and multicultural company
- Provide you with continuous challenge and room for learning
