Tech Blogs Digest 20.10 - 26.10
This week we AI-analysed 8245 posts for you, filtered out the chaff and hand-picked the wheat. Subscribe to stay up to date with future digests!
This week
🏗️ Architecture - AWS Bedrock for RAG, Kubernetes scheduler, and search scaling
🤖 LLMs in production - Agent monoliths, autonomous commerce, and a text-to-SQL chatbot
🛡️ Security - Fake AI recruiter, DevSecOps, and the AI handicap: attackers vs defenders
📊 Data - Delta sources for Spark
⚙️ DevOps - StatefulSet DNS, handcrafted Docker, and 30k deploys per day
💻 AI-assisted coding - AI-accelerated legacy system analysis and modernization
🧠 ML - Multi-size models with a single training process
🏗️ ARCHITECTURE
Building RAG application on AWS Using AWS Bedrock | 46 min read
Build a scalable RAG-powered chatbot on AWS - learn how AWS Bedrock streamlines ingestion, embeddings and inference for real-world AI systems
Deep Dive into the Kubernetes Scheduler Framework | 20 min read
Gain a clear view into how the Kubernetes scheduler framework orchestrates plugin lifecycles and decision-making to achieve high-scale, low-latency scheduling
From client-side filters to full-blown search infrastructure: the article maps out how to scale your search from zero to a million records - tech stacks, cost trade-offs and real metrics
🤖 LLMS IN PRODUCTION
See how turning one large AI into a team of specialist agents boosts reliability, cuts token waste, and drives smarter evolution
RAG vs Fine-Tuning: Choosing the Right Path | 8 min read
Which AI strategy fits you best - ground-truth retrieval or bespoke model tuning? Dive into the trade-offs between RAG and fine-tuning
I Gave Google’s Agent Payments Protocol (AP2) a Try: Here’s What I Learned About the Future of Autonomous Commerce | 15 min read
Google’s AP2 experiment hints at a future where AI handles your payments seamlessly - see what that could mean for finance
A hands-on guide to building a full-stack chatbot that converts plain English into SQL queries using LangChain, FastAPI, and Streamlit - turning everyday data questions into actionable results
From code-heavy workflows to intelligent automation, Snowflake’s new AI vision shows what true autonomous data engineering can look like
🛡️ SECURITY
A fake AI recruiter promises a dream job but hides dangerous malware that can take over your computer
See how a small team mastered Kubernetes security with DevSecOps—proving you don’t need a big crew to stay secure
Attackers thrive with tasks AI automates easily - defenders must rethink systems to counter this growing imbalance
📊 DATA
A deep-dive into the internals of Apache Spark Structured Streaming’s delta-table source reveals how offsets, micro-batches and the Delta Lake transaction log work together to enable efficient streaming
⚙️ DEVOPS
Why nats-headless service returns NXDOMAIN on EKS- A Practical Guide to Kubernetes Headless Service DNS | 7 min read
Learn how a small misconfiguration in service names or selectors can cause NATS headless DNS failures on Amazon EKS
See how the building blocks of containers: namespaces, pivot root, bridges, and more come together to recreate what Docker does under the hood
How can a company safely push 30,000 deployments per day? This article unpacks the automation, monitoring, and mindset behind the process at Mercado Libre
💻 AI-ASSISTED CODING
How LLMs Support Product Renovation: A Case Study | 19 min read
AI coding agents helped a team modernize legacy systems, speed up onboarding, and reveal hidden architecture insights in this real case study
🧠 ML
MatFormer: Train Once and Deploy Many | 17 min read
A single “train-once” model that can scale down dynamically: the article shows how MatFormer builds nested sub-models from one backbone to fit any deployment scenario


