Building Production ML Systems: MLOps Best Practices
Introduction
Machine learning in production is vastly different from notebooks โฆ
Machine learning in production is vastly different from notebooks โฆ
Fine-tuning large language models on custom data can be โฆ
When building production LLM applications, developers face a โฆ
Vector databases are the backbone of modern AI applications. They โฆ
Rust is increasingly becoming the language of choice for building โฆ
Rust’s ownership system is what makes it possible to โฆ
Tokio is Rust’s de facto standard async runtime, enabling โฆ
Unsafe Rust allows you to disable certain safety checks when โฆ
AWS cost optimization is one of the most underutilized ways to โฆ
Serverless is marketed as “pay-per-execution,” but many โฆ
Containerization (Docker) and orchestration (Kubernetes) are โฆ
Spot Instances are AWS’s ultra-discounted compute offering: โฆ
Privacy concerns in machine learning have become paramount as โฆ
Data science remains one of the most in-demand careers in tech. โฆ
Natural Language Processing (NLP) enables computers to understand, โฆ
Time series data is everywhereโfrom stock prices to sensor readings โฆ
Cloud security requires โฆ
Zero Trust replaces implicit trust โฆ
JWT is only one โฆ
The future of computing is distributed, and edge computing has โฆ
The cloud computing landscape has evolved dramatically. โฆ
APIs are the backbone of modern applications, enabling โฆ
Compute resources represent a significant portion of cloud spending โฆ
WebSockets enable bi-directional, real-time communication between โฆ
Node.js is ideal for building RESTful APIs. Its event-driven, โฆ
APIs are the connective tissue of modern software. From mobile apps โฆ
Building an AI API is different from traditional APIs. You deal โฆ
The era of cloud-dependent mobile AI is ending. Modern smartphones โฆ
Users expect mobile apps to be instant, smooth, and efficient. In โฆ
Mobile app privacy and security have become critical concerns in โฆ
Mobile development offers multiple paths: native iOS, native โฆ
Certificate revocation is a critical component of PKI security. โฆ
Email remains one of the most critical communication channels for โฆ
AMQP (Advanced Message Queuing Protocol) is an open-standard โฆ
API gateways have become the cornerstone of modern microservices โฆ
Ruby 2.0 introduced prepend as a new way to mix modules into classes. While include adds a module’s methods after the class in the โฆ
Ruby has a rich ecosystem of technical writing. These are the articles and resources that provide the deepest understanding of โฆ
A singleton method is a method defined on a specific object rather than on its class. Only that one object has the method โ no other โฆ
Dynamic dispatch is the ability to call a method by name at runtime, rather than hardcoding the method call at compile time. In Ruby, โฆ
In Ruby, instance variables (@name, @age) are private by default โ they can’t be accessed from outside the class. To expose them, โฆ
One of the most common uses of class variables (@@var) is tracking shared state across all instances of a class โ like counting how many โฆ
When you call a method that doesn’t exist on a Ruby object, Ruby doesn’t immediately raise a NoMethodError. Instead, it calls โฆ
A class instance variable is an instance variable (@var) defined on the class object itself โ not on instances of the class. Because โฆ
Class variables in Ruby are prefixed with @@ and are shared across the class that defines them and all its subclasses. They’re one โฆ
When you include a module in Ruby, the module’s methods become instance methods of the including class. But what if you want the โฆ