Engineering With Java: Digest #83
👋 Java Devs! Welcome to this week’s edition — where JVMs run fast, memory leaks run faster, and Spring still keeps everyone guessing 🤯
👋 Java Devs! Welcome to this week’s addition! I hope you’re all doing great.
This week, we cover essential insights on:
📢 Get actionable Java/Spring Boot insights every week, from practical code tips to real-world use-case based interview questions.
Join 6500+ subscribers and level up your Spring & backend skills with hand-crafted content, no fluff.
First 100 paid subscribers will get the annual membership at $50/year forever that is ~ $4/mo ( 71 already converted to paid, 29 remaining)
Before we kick off, meme of the day 😊
🗒️ Articles Of The Week (8)
Building Distributed HTTP Sessions with Spring Session MongoDB
This article explains how to implement distributed HTTP sessions in Spring Boot using Spring Session with MongoDB as the backing store, allowing session data to be shared across multiple application instances. It highlights how this approach improves scalability and fault tolerance compared to in-memory sessions. The key takeaway is that externalizing session storage enables stateless, horizontally scalable architectures in modern Spring applications.
Thread-Safe Native Memory in Java: VarHandle Access Modes Explained
This article explores Java’s native memory access modes introduced via the Foreign Function and Memory API, explaining how developers can safely and efficiently interact with off-heap memory. It covers different access modes (like read/write constraints) that improve safety compared to traditional unsafe operations. The key takeaway is that modern Java provides safer, structured control over native memory without relying on low-level hacks.
Using Java for Developing Agentic AI Applications: The Enterprise-Ready Stack in 2026
This article discusses how Java can be effectively used to build agentic AI systems by integrating LLMs with structured application logic and enterprise tooling. It highlights benefits like strong typing, scalability, and ecosystem maturity for building reliable AI workflows. The key takeaway is that Java is well-suited for production-grade AI agents, especially in enterprise environments where robustness and maintainability matter.
Generated Columns and Computed Columns in SQL
This article explains generated (computed) columns in SQL, which automatically derive values from other columns to simplify queries and enforce consistency. It covers use cases like denormalization, indexing, and performance optimization. The key takeaway is that computed columns reduce redundancy and improve maintainability in database design.
AI Is Rewriting Enterprise Java’s Playbook – and Vice Versa
This article explains how AI is reshaping enterprise Java, with developers increasingly using AI tools to accelerate coding, testing, and modernization efforts. At the same time, Java’s strengths—stability, scalability, and strong tooling—make it a solid foundation for building production-grade AI systems. The key takeaway is that AI and Java are evolving together, with each reinforcing the other in enterprise development.
10 Java Performance Mistakes Senior Developers Still Make
This article outlines common Java performance mistakes even experienced developers make, such as excessive object creation, poor collection choices, inefficient I/O handling, and improper use of streams or concurrency. It emphasizes that small inefficiencies at scale can significantly impact performance. The key takeaway is that understanding JVM behavior, memory management, and choosing the right data structures and patterns is crucial for writing high-performance Java applications.
Semantic Search Is an Architecture Problem
Semantic search is an architecture problem because the real challenge isn’t embeddings—it’s deciding when to use semantic, keyword, hybrid retrieval, reranking, caching, and filters, and how to combine them reliably at scale. Most systems fail not from vector search itself but from poor system design choices around it (chunking, indexing, latency, observability, and tool selection). The key insight is that “meaning-based search” only works well when the surrounding architecture correctly controls context, cost, and failure modes rather than relying on vectors alone
Which Java Construct Should You Use? Let Change Drivers Decide
This article argues Java isn’t inherently bloated or memory-heavy anymore, and that most modern memory issues come from poor design rather than the JVM itself. It highlights improvements like ZGC, Shenandoah, compact object headers, container awareness, and smarter GC tuning. Overall, modern Java is memory-efficient, and performance depends more on architecture than the language runtime.
▶️ Videos of the week (3)
Spring Debugger New Power: Where Should I Click to Demystify Spring Boot Magic?
This talk uses Spring Debugger and Spring’s bean resolution rules to show how “correct-looking” assumptions can still fail at runtime due to hidden framework behavior like environment post-processors and bean naming conventions. Through Spring Boot puzzlers, it demonstrates how properties are resolved from multiple sources with strict precedence, and how tools like IntelliJ debugger can trace the exact origin of overridden values. It also highlights Spring’s bean definition phase where naming conflicts or conventions can change injection behavior in unexpected ways.
Java and Post-Quantum Cryptography
This talk explains post-quantum cryptography (PQC) and why future quantum computers could break today’s public-key encryption like RSA and ECC using Shor’s algorithm. It walks through NIST’s standardization of PQC algorithms (ML-KEM, ML-DSA, etc.), the urgency created by “store now, decrypt later” attacks, and the shift toward hybrid cryptographic schemes. It also highlights how Java is evolving with new APIs and TLS support to integrate PQC securely and ensure long-term protection.
Embabel Tools & MCP Servers: Supercharge Your Java AI Agents
This talk demonstrates building a Spring-based AI agent system using structured “actions”, tools, and MCP servers to generate a full blog-writing pipeline with draft, review, TLDR, and front-matter generation. It shows how LLM tools (like reading-time estimation) and external MCP web search can be injected to enrich outputs with real context. The key takeaway is that modern agent design focuses on composable, tool-augmented workflows rather than single prompts, improving reliability, structure, and control over LLM behavior.
🔥 Recently Published In-house Blogs (4)
Spring Data Interview Question - Identify & Optimize Slow SQL Queries
Spring Data Interview Question : Efficient Keyset Pagination with Spring Data WindowIterator
Java Bug Fix Interview Question - Retry Utility with Silent Failures
Java Interview Question - Detecting Duplicate Product Titles
Thats all for this week friends! Thanks for reading this far. If you liked it please share with your network.
Happy Coding 🚀
Suraj
Subscribe | Sponsor us | LinkedIn | Twitter