--- title: "Shlomo Friman (@shlomofr) | Hashnode" description: "Exploring the intersection of software archaeology and modern enterprise intelligence. Read the latest articles by Shlomo Friman on Hashnode." url: "https://hashnode.com/@shlomofr" --- @shlomofr # Shlomo Friman @shlomofr · Austin, USA · Joined April 2026 Exploring the intersection of software archaeology and modern enterprise intelligence. [Website](https://www.in-com.com/) ## Comments [One thing I'd add from the enterprise side, where codebases tend to be measured in decades rather than years: the golden files principle runs into a wall when the "approved pattern" in the codebase is itself a workaround for something nobody documented. I've seen teams point AI to golden files that were golden by convention, not by design — the file was just the one nobody broke. The AI learns the workaround as the standard, and the instructions quietly encode technical debt as policy.](https://hashnode.com/posts/code-generation-and-review-with-ai-in-complex-repositories/6a0207c3fca21b0d4b49990a/comment/6a045e385418a55a522d156d) Comment · Article · May 13 · [Code Generation and Review with AI in Complex Repositories](https://hashnode.com/posts/code-generation-and-review-with-ai-in-complex-repositories/6a0207c3fca21b0d4b49990a) [The CI/CD boundary problem is real and most teams solve it with duct tape inside their existing pipeline config. Lightweight, self-hosted, Go — the architecture choices make sense for something that sits in the critical path of every release. Will follow the progress.](https://hashnode.com/forums/thread/i-built-a-react-component-playground-where-you-own-every-generated-file/comment/69f46406909e64ad07a1e455) Comment · Thread · May 1 · [I built a React component playground where you own every generated file](https://hashnode.com/forums/thread/i-built-a-react-component-playground-where-you-own-every-generated-file) [Solid write-up. The network\_mode: service:gluetun pattern for routing dependent containers is clean and underused. One caveat worth adding: any ports those downstream containers need have to be declared on the Gluetun container, not on themselves. Catches a lot of people the first time.](https://hashnode.com/posts/setting-up-wireguard-on-synology-dsm-7-using-docker-and-gluetun/69f42d8c5d047222f61f4e05/comment/69f46341909e64ad07a14d20) Comment · Article · May 1 · [Setting up WireGuard on Synology DSM 7 using Docker and Gluetun](https://hashnode.com/posts/setting-up-wireguard-on-synology-dsm-7-using-docker-and-gluetun/69f42d8c5d047222f61f4e05) [The six-versus-thousands framing is the most honest description of microservice operational overhead I have read. The AI triage pipeline is the right answer. Would love a follow-up post once the testing automation is live.](https://hashnode.com/posts/managing-security-issues-with-microservices/69f38420909e64ad07fbb579/comment/69f46299909e64ad07a0d672) Comment · Article · May 1 · [Managing Security Issues with Microservices](https://hashnode.com/posts/managing-security-issues-with-microservices/69f38420909e64ad07fbb579) ["Can this actually survive production?" is the right question and most vendors cannot answer it honestly because their demo environment has none of the edge cases that production has. Clean inputs, cooperative APIs, predictable load. Ship it into the real world and all three assumptions break within the first week.](https://hashnode.com/forums/thread/truth-hurts-why-do-most-enterprise-ai-automation-workflows-still-break/comment/69f34329909e64ad07c0cb17) Comment · Thread · Apr 30 · [TRUTH HURTS! Why Do Most Enterprise AI Automation Workflows Still Break?](https://hashnode.com/forums/thread/truth-hurts-why-do-most-enterprise-ai-automation-workflows-still-break) [Mostly agree, but "done over perfect" is worth being careful with. In my experience junior developers hear that and use it to justify shipping things that are not actually done — they are just finished enough to close the ticket. The senior version of that principle is closer to: know precisely which corners you are cutting and why, and make sure the team knows too. Cutting corners invisibly is how you accumulate the kind of technical debt that nobody can explain when you try to fix it two years later.](https://hashnode.com/forums/thread/the-senior-dev-secret-it-s-rarely-about-the-code/comment/69f3130d909e64ad078edb64) Comment · Thread · Apr 30 · [The "Senior" dev secret: It’s rarely about the code.](https://hashnode.com/forums/thread/the-senior-dev-secret-it-s-rarely-about-the-code) [The no-code angle here is underrated. Most RAG tutorials assume you are building everything from scratch with Python and a vector database, which puts the approach out of reach for teams that live in the Microsoft ecosystem. Being able to point Azure AI Search at OneLake through the portal and have chunking, embedding, and indexing handled as a managed service dramatically lowers the barrier. For enterprise teams already invested in Fabric this is a much more realistic starting point than spinning up a custom LangChain pipeline. Good practical post.](https://hashnode.com/posts/azure-ai-search-for-fabric-one-lake-unstructured-data/69c2d1f492029d915f605254/comment/69f3127e909e64ad078e4d9b) Comment · Article · Apr 30 · 1 · [Azure AI search for Fabric One Lake unstructured data](https://hashnode.com/posts/azure-ai-search-for-fabric-one-lake-unstructured-data/69c2d1f492029d915f605254) [The hybrid model you describe is the right direction but the discipline problem you mention is actually the whole problem. In practice, the boundary between "routine task" and "complex architectural question" is almost never clear in the moment. You start a small refactor, it surfaces a design issue, and you are suddenly in an expensive API session you did not plan for. We tried the same approach on our team and gave up on the split after two months: the categorization overhead was its own friction. What actually worked was setting a hard monthly cap on API spend and letting it force natural rationing, rather than trying to classify tasks upfront.](https://hashnode.com/posts/the-thing-everyone-gets-wrong-about-ai-coding-costs-including-me/69f2d46d0982df87b6148dd6/comment/69f311ce909e64ad078db115) Comment · Article · Apr 30 · [The Thing Everyone Gets Wrong About AI Coding Costs (Including Me)](https://hashnode.com/posts/the-thing-everyone-gets-wrong-about-ai-coding-costs-including-me/69f2d46d0982df87b6148dd6) [The traffic manager analogy for the Event Loop is a good one. I usually explain it to new engineers as a restaurant kitchen with one chef: the chef does not stand at the oven watching a dish cook — they hand it off, move to the next order, and come back when it is ready. The moment the chef starts watching the oven is when the whole kitchen backs up. Same thing happens when you block the Event Loop. Nice post, solid foundation for anyone learning Node internals.](https://hashnode.com/posts/the-node-js-event-loop-explained/69f1ae0cad6e531c72790b2c/comment/69f30fb9909e64ad078a43f3) Comment · Article · Apr 30 · 2 · [The Node.js Event Loop Explained](https://hashnode.com/posts/the-node-js-event-loop-explained/69f1ae0cad6e531c72790b2c) [Claude is better when you want clarity. ChatGPT is better when you want capability. Claude feels like a thoughtful writer. ChatGPT feels like a toolchain. If you’re doing serious work, you don’t pick one, you route tasks between them.](https://hashnode.com/forums/thread/is-claude-better-than-chatgpt/comment/69f30eca909e64ad0789521d) Comment · Thread · Apr 30 · 1 · [Is Claude better than ChatGPT?](https://hashnode.com/forums/thread/is-claude-better-than-chatgpt)