Overview
Discover how machine learning and Azure combat illegal fishing in this 46-minute conference talk. Learn to construct a cloud-first architecture using serverless and data analytics technologies to detect and tackle this global issue. Explore the principles and challenges of designing such solutions, and see how the architecture provides cloud benefits like reliability, scalability, and security while keeping compute costs under £10 per month. Gain insights into protecting oceans, organizational landscapes, cloud evolution, commodity computing, and serverless characteristics. Delve into developing feature pipelines, data projections, serverless machine learning and inference, experimental methods, proximity algorithms, and orchestration with Durable Functions. Discover performance optimization techniques in .NET, including low allocation compute and list pre-allocation. Walk away with valuable takeaways and references for further exploration.
Syllabus
Intro
What will learn?
Protecting the world's oceans
The organisational landscape
Cloud evolution
Commodity computing
Cost comparison
Serverless characteristics
Developing a feature pipeline
Data projections
Serverless machine learning
Serverless inference
Experimental method
Proximity algorithm
Orchestration
Durable Functions
Performance in NET
Techniques for low allocation compute
List pre-allocation
Reference vs Value types
Takeaways
References
Taught by
NDC Conferences