Yasir M1
1QCIF, Brisbane, Australia
Biography:
Muhammad Yasir is a Software Developer at QCIF, working on data platforms and cloud-based systems for wildlife image processing and research workflows. He has experience in full-stack development, cloud infrastructure, databases, and scalable data processing systems. His work focuses on building reliable software solutions that support research, data management, and digital transformation.
Abstract:
Australia’s biodiversity challenges require research infrastructure capable of supporting vast and rapidly growing wildlife monitoring datasets. Camera-trap projects generate large numbers of images and metadata across multiple locations using different approaches. Turning this data into actionable insights requires more than storage, it demands practical decisions regarding data organisation, processing workflows, usability, collaboration, and long-term reuse.
Using the Wildlife Observatory of Australia (WildObs) as a developer-focused case study, this presentation explores the design and implementation of its underlying wildlife image platform. We first examine the image platform from the user perspective by demonstrating how users can create projects, upload camera-trap metadata, organise deployments, manage images, and trigger computer vision species detection, enabling seamless access for researchers, environmental managers, government agencies, and non-technical contributors alike.
We will then dive behind the interface to examine the engineering decisions required to operate such a system at scale. We will cover core platform architecture, the trade-off between monolithic and microservice approaches, alongside frontend and backend technologies, cloud infrastructure for computationally intensive jobs, scalable image storage, queue-based workflows, metadata validation, access control, monitoring, and deployment practices. It also discusses how computer vision models are integrated to support automated species identification by running detection models, displaying confidence scores for human review and supporting data exports for further research.
The core focus is on how technical design choices impact usability, reliability, and scientific trust. The presentation concludes with practical lessons for research software engineers and infrastructure teams in building scalable, sustainable, and user-centered research platforms.
