Oleg Burov

🔥 Platform threads were never built to scale. We just pretended they were. 1MB stacks. OS scheduling. RejectedExecutionException at 3 AM. Java made us duct-tape our way through concurrency. Thread pools weren’t enough. Reactive made things harder. We weren’t scaling — we were surviving. 🚨 In Thread Wars: Episode 1, I break down why thread-per-request hit a wall, how async became a necessary evil, and what we were really trying to solve. 🧵 Next up: virtual threads — not just a new API, but a shot at reclaiming sanity in backend Java. 📖 Read Episode 1 here: https://lnkd.in/g7UPrD3Y #Java #VirtualThreads #ThreadWars #BackendEngineering #JavaConcurrency #ScalableJava #ReactiveFatigue #Java21 #CleanConcurrency #SystemDesign

5

Venting a little frustration here while developing a new mapping feature in RidgeText. Gemini is much lazier than Claude. It actively finds ways to avoid doing work, like saying "but the direct REST API is preferred for our current architecture" instead of determining why it's preferred when our structure doesn't matter. We're an LLM tool using Vertex functions. Whether those functions are REST, server-side scripts, MCP calls, or anything else doesn't matter. I hate that I have to tell Gemini to do better when Claude consistently lays out proper research plans. https://ridgetext.com

4

2 Comments

In a new blog post, Marc Bowes looks at how updates work in Aurora DSQL, and what it means for scalable schema design. Read it here: https://lnkd.in/gmGEK8Dv A couple take-aways for application builders: - You mostly don't have to worry about hot read keys in DSQL, even in read-write transactions. DSQL can scale out per-key read throughput, and the design requires no co-ordination between transactions reading the same key. (Thank you, MVCC and Time Sync!) - Scalable applications do need to avoid high-contention write keys when doing updates. Writes need to be coordinated and ordered (to implement the I in ACID), which limits per-key throughput. - Most applications can avoid hot write keys with the right schema design. Most applications have already been written this way, because high-contention keys perform poorly on relational databases of all kinds.

240

11 Comments

I wish more companies were like Salesforce in offering a sandbox env for development and testing. That said, they have to for it because nothing is straightforward when integrating vendor offerings into their application.

2

While everyone was talking about Opus 4.6, for me the true killer feature of the recent Claude Code updates is the agent teams. It's been something i've been trying to achieve with customization for a while. Custom agents, orchestration scripts, specific CLAUDE.md instructions to coordinate work... with some degree of success. But what Anthropic shipped natively is a WHOLE different level. What makes this stand out is the inter-agent communication. We're not talking about simple fan-out/fan-in where you spawn workers and collect results. These agents talk to each other. Peer-to-peer messaging, dependency-aware task graphs that auto-unblock, agents that self-claim work from a shared task list. The lead can even enter Delegate Mode where it does ZERO implementation, only coordination. The image below is from one of my setups. A team manager orchestrating a librarian agent, a PhD lead, and 5 research sub-tasks with blocking dependencies. The librarian unblocks the research tasks, the PhD lead aggregates everything. All coordinated autonomously. And with this a whole new world of orchestration just unveiled itself. Distributing work across agents is the "easy" part. You break down tasks, assign owners, define dependencies. The HARD part, and what MOST stands out now, is aggregation. How do you take the output of 5 parallel agents, each with their own context window, and synthesize it into something coherent? That's the new skill. Anthropic themselves used 16 parallel agents to build a 100,000 line Rust C compiler that compiles the Linux 6.9 kernel. No human actively coding. ~$20,000 in API costs over ~2,000 sessions. We went from pair-programming with AI to managing AI engineering teams. The skills that transfer are the ones from engineering management: task decomposition, context management, knowing when to intervene vs let the team self-organize. This is a new paradigm, and i think it opens up several possibilities we haven't fully explored yet. ref.: https://lnkd.in/dVCe344z

68

12 Comments

Agents are the new apps. And they just went multiplayer. The software engineering landscape is undergoing a seismic shift. Google just dropped Firebase Studio. And it might just replace half the AI builder startups out there. You’ve probably seen tools like Lovable, bolt.new, Vercel's v0, or Cursor. They let you build AI-powered apps without writing much code. Describe what you want — boom — they create UIs, workflows, and automation for you. It’s fast. It’s smart. It’s changing how software is built. Now imagine that… backed by Google. What is Firebase Studio? Firebase Studio is Google’s new end-to-end app builder for AI-first products. It combines: ✅ UI generation ✅ Agent workflows (like AgentSpace) ✅ Firebase backend (auth, storage, hosting, etc.) ✅ Google’s best Gemini models AgentSpace is like a command center for AI agents. It lets you create multistep workflows, connect APIs, and manage how agents behave. Great for product teams prototyping internal tools, automations, or AI assistants. But it’s still pretty early-stage. Right now, Firebase Studio and A2A (Agent-to-Agent) are two separate moves from Google — but they are strategically aligned, and integration is very likely coming. You need to manage agents, instructions, tools, and sometimes code. Why this matters for product managers? → The barrier to building software is decreasing. → With Firebase Studio, a product manager can do what used to take a team. The AI agent is becoming the new app. Not just chat. Agents that do stuff: send emails, fetch data, run logic. Google wants to standardise agents across platforms. Their new Agent-to-Agent protocol (A2A) means your agents could soon talk to each other — across apps, tools, maybe even companies. Enter MCP: The USB-C for AI Anthropic’s Model Context Protocol (MCP) is an open standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools. MCP is a standardised protocol designed to streamline the way AI models communicate with external systems, eliminating the need for custom integrations. With so many advancements happening rapidly, it’s challenging to keep up and navigate them within the realities of development. Yet, it has never been more exciting to be in the software engineering space. #AI #FirebaseStudio #ProductManagement #Innovation #TechTrends #MCP #A2A

16

1 Comment

Every PMM team needs to do this now! Get access to the source code repo. Here is why: 1. Ask questions in Claude or Cursor 2. Uncover how the product works 3. Answer technical and high level questions quick 4. Format outputs by audience 5. Even draw flow diagrams Now there is never a new question you can’t answer in minutes. No need to ask a PM. No need to wait.

12

Here’s the uncomfortable truth: Most teams don’t skip tests. They postpone them because testing slows momentum. McKenna Barlow just shared how GitHub Copilot Testing for .NET is now GA in Visual Studio 2026 removes that excuse as it’s engineered specifically for unit testing, not generic code generation. This isn’t a prompt ➡️ response toy. It’s an end-to-end test workflow embedded in your IDE. This is what “agentic development” is supposed to look like. The real shift? Testing stops being a chore and starts becoming default behavior. Will this change how your team approaches unit tests, or just make shipping faster? Read the full breakdown here: https://lnkd.in/g3UJzD7J #dotnet #aiengineering #visualstudio #unittesting #dontbeboring

29

At VeteranCrowd, every API contract starts with a Zod schema. Not a Swagger doc. Not a TypeScript interface. A runtime-validated schema that generates types, validates requests, and documents itself. This isn't just preference; it's architectural discipline. When your Lambda receives a merchant onboarding payload, SMOZ middleware validates it against the schema before your handler ever runs. Invalid requests never touch business logic. Error messages are auto-generated from the schema structure. The payoff compounds: - New endpoints inherit validation for free - Schema changes propagate to types, docs, and tests automatically - Integration tests validate against the same schemas as production - Onboarding new developers means pointing them at the schema, not a wiki We apply this pattern everywhere now: jeeves-watcher config (Zod + JsonMap), STAN context files, even social post generation. Schema-first isn't a tool choice. It's a quality multiplier.

1

1 Comment

I see this question come up a lot: MariaDB or something else once scale becomes a problem? This blog breaks down a real-world comparison between MariaDB and TiDB, covering what changes operationally when you need horizontal scale, online schema changes, and strong consistency. If you’re running MariaDB today and thinking about what comes next, this is a helpful, practical read.👇 https://ow.ly/NpxB50Y7M96 #MariaDB #MySQL #DatabaseArchitecture #DistributedSQL #TiDB

10

"Our Architecture Is Modern and Scalable" There are three sentences that appear in every healthcare IT acquisition deck. Your deal team has seen them a hundred times. They nod along because the phrases sound reasonable. They're not reasonable. They're warning signs. This is the first. "Our architecture is modern and scalable." What they mean: We rewrote the front end in React sometime in the last five years. What's hiding underneath: The backend is a monolith running on the same bones since the Obama administration. The database schema was designed when the company was twelve people. "Scaling" means throwing hardware at it and praying. I've walked into systems where "modern" meant the UI looked clean while the guts were held together with duct tape and tribal knowledge. Pretty dashboards. Terrifying infrastructure. The real cost? You discover post-close that you can't integrate without a rebuild. That "straightforward" migration becomes an 18-month odyssey. The synergies in your model evaporate while your team fights fires nobody saw coming. Here's what I ask instead: Show me what happens when you 2x your transaction volume tomorrow. Not theoretically. Actually show me. The answer tells you everything the deck won't.

20

1 Comment

This matches my experience building for clients as well. We had to remove steps from old pipeline to let the model 'decide' the best. It worked better than our hard coded workflow. Thinking image models have changed the game. Thinking is now removed from the workflow. for part of workflow - which was earlier newded. Result? 1. better outputs 2. faster and cheaper AI runs. There are two important takeaway from this: 1. also applicable to images 2. the 'intelligence'of what the next step is was earlier in workflowz -> it can be easily in the models now (model + tools / skills)

2

I joined Tim te Beek and Sam Snyder on Code Remix Weekly last week to talk about code search and why “searching text” often isn’t the same as “searching code.” A few highlights from our chat:  • Traditional search can return an “avalanche of irrelevant results” when names overlap (think 𝚂𝚝𝚛𝚒𝚗𝚐𝚄𝚝𝚒𝚕𝚜, 𝙻𝚘𝚐𝚐𝚎𝚛, etc.)  • The opposite problem is missing matches when the thing you care about is “buried” (types, inheritance, transitive dependencies, etc.)  • IDEs do have semantic context, but they’re limited to what you have open We looked at several examples, from type and method usage, sensitive API endpoints, and dependency insight, to secrets scanning, SBOM generation, and deprecated usage. Watch here: https://lnkd.in/gBiSS3fH

17

PTC exits from #IIoT business selling, in one shot both ThingWorx, a PTC Technology and Kepware, a PTC Technology. The biggest IT vendor jumped in the shop floor years ago with two acquisitions now is selling back all to PE. The IoT platforms were dead years ago with increasing capabilities of analytics tools (like SAS, to mention just our partner in KME project ) and no-code #IoT integration platform as our Alleantia. The #iot today is just a different data domain without needs of special platform.

11

If you’re planning to modernize a legacy system that's been running for years and still manages a high volume of client traffic, the strangler fig pattern might be something that can help. Read Martin Fowler's blog post about the Strangler Fig pattern application. https://lnkd.in/gxKMuQ5G https://lnkd.in/gMxK3KUS If you’ve walked this path, what tipped your decision one way or another? #modernization #StranglerFig #SRE #softwarearchitecture #engineeringmanagement

40

So much this! Although I'd say the best replacement for JSON for configs is not a YAML, but some trivial language. Too bad we do not have good self-contained languages, so that we can trust it that the "config code" does not have side effects and is overall predictable. Otherwise Python or even Golang / Rust would do. I personally am a huge fan of GCL [1] btw! Can write map/reduce "code", and yet the code itself consumes the result of this "config file" as if it's a static JSON object of a certain hierarchy.

7

2 Comments

🚀 Big upgrade for real-time AI. With the GA of Zerobus Ingest (part of Lakeflow Connect), Databricks just made streaming architectures dramatically simpler. No brokers. No message bus ops. Direct, serverless ingestion into governed Delta tables. For telecom companies, this is massive: 📡 Ingest network telemetry in real time ⚡ Power AI models for churn, SLA breaches, anomaly detection 💰 Reduce Kafka/Event Hub infrastructure costs 🔐 Maintain governance natively in the Lakehouse This isn’t just a feature release — it’s Databricks collapsing the streaming stack into the platform itself. Fewer moving parts. Lower latency. Stronger AI systems. The lakehouse keeps getting better. 🔥 Ali Ghodsi Rajesh Sreepada Tushar Sinha Diptesh Singh Matei Zaharia Amit Singh #Cognizant #AIA #Databricks #DatabricksMVP #AI #Telecom #Streaming #Lakehouse #DataEngineering

6

Our partners at CREATEQ are spot on: "Agents are the new apps. And they just went multiplayer." The software engineering landscape is in the midst of a rapid and profound seismic shift, powerfully evidenced by Google's recent unveiling of Firebase Studio. Now more than ever, for everyone in the software engineering industry, understanding and proactively adapting strategies to these changes is absolutely crucial. Don't get left behind in this exciting evolution! Thank you Marc Gasser for sharing this vital insight! #TechTrends #AI #FirebaseStudio #AdaptOrEvolve #Createq

3