Duties & Responsibilities
1. Application Development and AI Systems Engineering
- Design and deploy computer vision pipelines using advanced architectures (CNNs, Transformers, YOLO) to solve retail challenges like inventory tracking, people counting, and on-shelf availability.
- Optimize model inference for diverse environments, including handheld terminals (HHTs), edge devices, and cloud-based CCTV integration (e.g., Dahua HTTP APIs).
- Implement image preprocessing and augmentation techniques to ensure model robustness against varying store lighting and camera angles.
- Continuously research and implement state-of-the-art ML/AI methodologies to ensure optimal balance between predictive accuracy, interpretability, and business relevance.
2. Scalable API Development and System Architecture
- Develop high-performance Python APIs (FastAPI/Flask) to integrate AI capabilities into retail mobile and web applications.
- Implement concurrency and parallel computing (multi-threading, AsyncIO, GPU acceleration) to handle high-throughput data from hundreds of retail outlets.
- Utilize advanced deep learning architectures (CNNs, RNNs, transformers) to analyze unstructured data such as images, text, and sequential signals
3. Cloud Infrastructure and AWS Integration
- Architect and maintain scalable AI solutions using AWS cloud services (Sagemaker, Lambda, EC2, S3) to support large-scale retail operations.
- Manage cloud resource utilization to balance computational performance with cost-efficiency across 500+ store locations.
4. MLOps, Dockerization, and CI/CD
- Integrate DevSecOps best practices by embedding security scanning, identity management (IAM), and data encryption into the application development lifecycle.
- Ensure the maintainability and reproducibility of the codebase through rigorous version control (Git) and automated documentation.
- Collaborate with IT and Security teams to ensure AI applications comply with corporate data governance and retail privacy standards.
- Continuously track model performance and implement updates to maintain accuracy, fairness, and business relevance. (e.g., Docker, CI/CD, Airflow, MLflow)
5. DevSecOps and Lifecycle Management
- Integrate DevSecOps best practices by embedding security scanning, identity management (IAM), and data encryption into the application development lifecycle.
- Collaborate with IT and Security teams to ensure AI applications comply with corporate data governance and retail privacy standards.
6. Business Impact, Maintenance, and Innovation
- Partner with store operations and product teams to translate retail pain points into ""smart"" automated solutions that reduce manual labor and improve efficiency.
- Provide end-to-end maintenance for deployed AI applications, ensuring high availability and rapid troubleshooting of production issues.
- Lead rapid prototyping sprints to validate emerging technologies (e.g., Generative AI, Agentic workflows) and evaluate their potential for driving measurable ROI.