Chapters: Ch 1 · Ch 2 · Ch 3 · Ch 4
APIs & GraphQL
Why, after 6 years, I’m over GraphQL (bessey.dev) https://bessey.dev/
A candid retrospective by Phil Bessey after six years building with GraphQL in production. Concludes that GraphQL’s complexity, N+1 problems, caching difficulties, and security surface area outweigh its benefits for most applications compared to well-designed REST APIs. [→ data-engineering]
Conexus data interoperability
A note on Conexus — a data interoperability framework or standard for connecting heterogeneous data sources. Likely referenced in the context of building data pipelines that bridge different schema formats or protocols. [→ data-engineering]
Scheduling
Artificial Intelligence to Solve Production Scheduling Problems in Real Industrial Settings Systematic Literature Review — particle swarm optimization, neural networks, reinforcement learning are the most widely used techniques. AI solutions have reduced production costs, increased energy efficiency, improved scheduling. www.mdpi.com https://www.mdpi.com/2079-9292/12/23/4732
A systematic literature review in MDPI Electronics surveying AI approaches to production scheduling in industry. Finds that particle swarm optimisation, neural networks, and reinforcement learning dominate the literature, with documented results showing cost reduction, energy savings, and throughput improvements in real factory settings. [→ data-engineering; machine-learning-ai; algorithms-data-structures]
Graph neural networks for exact solving scheduling problems (local PDF reference) file:///Users/vishwas/Downloads/2022Exactsolvingschedulingproblemsacceleratedbygraphneuralnetworks_RG.pdf
A 2022 research paper on using GNNs to accelerate exact branch-and-bound solvers for scheduling problems. The GNN learns branching heuristics from problem structure, dramatically reducing the search space while preserving optimality guarantees — a hybrid of ML and classical OR. [→ algorithms-data-structures; machine-learning-ai]