Links tagged “architecture”
39 links, newest first.
earendil.com
inference APIs make session state non-portable, and that is lock-in.
timeseriesofindia.com
UPI: Anatomy of a Transaction - one Indian instant payment followed layer by layer, seven actors between the PIN and the result.
martinfowler.com
Building reliable agentic AI systems: a production write-up of making LLM agents dependable, on Fowler's site.
overreacted.io
There are no instances in ATProto: Dan Abramov on the data model behind Bluesky, where your identity and posts are portable instead of tied to one server. A clean look at designing a protocol for exit.
dbos.dev
Postgres as a backend for async tasks.
github.com
table: A Periodic Table of System Design Principles
devops.com
What Happens When AI Starts Writing Your APIs?: how AI's ability to quickly generate APIs is creating a new class of security risks and technical debt.
towardsdatascience.com
How to Ensure Reliability in LLM Applications: An article by Eivind Kjosbakken discusses strategies for ensuring robustness in LLM implementations.
nordicapis.com
API Architectural Styles of 2025: A nice summary of API options available nowadays; nothing innovative here but can be a good refresher.
florian-kraemer.net
Most RESTful APIs aren't really RESTful
stripe.com
A few years ago, Bloomberg Businessweek published a feature story on Stripe.
rabbitmq.com
AMQP 0-9-1 Model Explained This guide provides an overview of the AMQP 0-9-1 protocol, one of the protocols supported by RabbitMQ.
rabbitmq.com
Exchange to Exchange Bindings In AMQP 0-9-1 the queue.bind method binds a queue to an exchange so that messages flow (subject to various criteria) from the exchange (the source ) to the queue (the destination ).
skillachie.com
How exchange-to-exchange bindings route messages between RabbitMQ exchanges, and where they beat binding straight to a queue.
jack-vanlightly.com
So the implication is that you need at least as many partitions as the most scaled out consumer.
medium.com
Beyond Interactive: Notebook Innovation at Netflix By Michelle Ufford, M Pacer, Matthew Seal, and Kyle Kelley Notebooks have rapidly grown in popularity among data scientists to become the de facto standard for quick prototyping and exploratory analysis.
vorpus.org
Every concurrency API needs a way to run code concurrently.
htdp.org
Please send reports about mistakes to matthias @ ccs.neu.edu
philcalcado.com
Since their first introduction many decades ago, we learnt that distributed systems enable use cases we couldn’t even think about before them, but they also introduce all sorts of new issues.
