https://pinishv.com/Recent content in PiniShv on PiniShv | AI, DevOps & Engineering LeadershipHugo -- gohugo.ioen© 2026 Pini Shvartsmanhttps://pinishv.com/articles/one-man-show-company/Wed, 29 Apr 2026 22:00:00 +0300https://pinishv.com/articles/one-man-show-company/A company used to start with people. Now it can start with one person and a swarm of AI agents that draft, build, test, sell, and support faster than any team you could hire. Most founders will turn this into a vending machine for bankruptcy. The Kolboynik who learns to manage agent labor, not just use AI, gets a real shot. Three buckets of risk, six starter agents, nine non-negotiable safety rules, and the brutal question that separates operators from victims.https://pinishv.com/articles/eu-ai-act-100-days-engineering-not-started/Fri, 24 Apr 2026 16:00:00 +0300https://pinishv.com/articles/eu-ai-act-100-days-engineering-not-started/August 2, 2026 is the enforcement deadline for EU AI Act high-risk obligations. From today, that’s exactly 100 days. In most orgs, the legal team is tracking this and the engineering team hasn’t been formally told what they need to ship. By July that gap will not be recoverable. Here’s what Articles 5, 12, 14, and 50 actually require when you translate them into code, and a 100-day plan to ship on time.https://pinishv.com/articles/vibe-coding-backlash-seniors-lose-argument/Fri, 24 Apr 2026 14:00:00 +0300https://pinishv.com/articles/vibe-coding-backlash-seniors-lose-argument/Forbes just said vibe coding will break your company. Senior engineers are organizing against it. The data is on their side: independent audits keep finding materially more issues in AI-co-authored code, no-code AI platforms are shipping apps with real security holes, and a Replit agent deleted a live production database during a code freeze last summer. Seniors are still about to lose the argument in every quarterly review unless they can make their judgment legible. Here’s what actually needs to ship.https://pinishv.com/articles/end-of-courses-learn-from-ai-like-a-toddler/Fri, 24 Apr 2026 10:00:00 +0300https://pinishv.com/articles/end-of-courses-learn-from-ai-like-a-toddler/Remember when shipping an app meant 40 hours of video courses and weeks of syntax memorization? An agent builds it in three minutes now. The 40-hour prerequisite is dead; targeted, just-in-time learning is more valuable than ever. You now have two choices: become a prompt-runner any motivated middle-schooler can replace, or become the Kolboynik architect who learns from every agent output the way a toddler learns to speak. Slower code path, faster growth curve.https://pinishv.com/articles/anthropic-q1-2026-catching-the-wave/Mon, 13 Apr 2026 10:00:00 +0300https://pinishv.com/articles/anthropic-q1-2026-catching-the-wave/Anthropic shipped 120+ features in 90 days, then blocked OpenClaw from using Claude subscriptions. The same company building the best developer tools in AI is also building walls around them. I’ve always spread my bets across providers—but when one company moves this fast, even diversification has a cost.https://pinishv.com/articles/ide-becoming-mission-control/Sun, 05 Apr 2026 10:00:00 +0200https://pinishv.com/articles/ide-becoming-mission-control/Cursor 3 rebuilt its UI around agents. GitHub calls Agent HQ ‘mission control.’ VS Code is ‘your home for multi-agent development.’ JetBrains Air says the quiet part out loud: build tools around the agent, not the editor. The file tree isn’t disappearing. It’s just no longer the main character.https://pinishv.com/articles/local-llms-your-stack-is-rented/Sat, 04 Apr 2026 18:00:00 +0200https://pinishv.com/articles/local-llms-your-stack-is-rented/Anthropic just told Claude Code users that third-party harnesses need separate billing. Google dropped Gemma 4 under Apache 2.0 across phone-to-workstation tiers. One story is about dependence. The other is about escape velocity. The local LLM landscape finally crossed from ‘cute demo’ to ‘actually useful.’https://pinishv.com/articles/ai-code-cheap-to-produce-not-to-own/Thu, 02 Apr 2026 12:00:00 +0200https://pinishv.com/articles/ai-code-cheap-to-produce-not-to-own/AI accounts for 42% of committed code. 96% of developers don’t fully trust the output. Only 48% always verify before committing. The gap between how fast we generate code and how well we govern it is the real risk of AI-assisted development.https://pinishv.com/articles/shadow-ai-most-dangerous-sentence/Thu, 02 Apr 2026 10:00:00 +0200https://pinishv.com/articles/shadow-ai-most-dangerous-sentence/35% of developers access AI coding tools through personal accounts. AI lets one person bypass every paved road the organization built, very fast and very quietly. Shadow AI isn’t about rogue employees. It’s about productive people touching systems the company is responsible for.https://pinishv.com/articles/claude-code-leak-why-it-matters/Thu, 02 Apr 2026 08:00:00 +0200https://pinishv.com/articles/claude-code-leak-why-it-matters/Anthropic’s Claude Code accidentally shipped internal source code in a release. Not a breach. A packaging mistake. A missed step. That’s exactly the kind of failure AI makes more likely, because the dopamine is in generating the feature, not in validating the artifact that ships.https://pinishv.com/articles/cisco-llm-security-leaderboard/Thu, 26 Mar 2026 10:00:00 +0200https://pinishv.com/articles/cisco-llm-security-leaderboard/Cisco just published a public leaderboard scoring LLMs on how well they resist attacks. Anthropic dominates the top 10. Multi-turn attacks are where most models crack. The rankings are interesting, but the real value is the question they force every engineering team to ask.https://pinishv.com/articles/openai-kills-sora-focus-enterprise/Wed, 25 Mar 2026 08:00:00 +0200https://pinishv.com/articles/openai-kills-sora-focus-enterprise/OpenAI shut down Sora, killed a $1 billion Disney deal, and pivoted to enterprise and robotics. The most-downloaded AI video app on iOS, gone. This isn’t a product failure. It’s a strategic signal about what actually makes money in AI.https://pinishv.com/articles/claude-computer-use-dispatch/Mon, 23 Mar 2026 14:00:00 +0200https://pinishv.com/articles/claude-computer-use-dispatch/Anthropic just shipped computer use for Claude. It can click, scroll, navigate your browser, open files, run dev tools, and submit PRs. Pair it with Dispatch and you can assign tasks from your phone while Claude works on your Mac. This is the jump from ‘AI that talks’ to ‘AI that does.’https://pinishv.com/articles/deerflow-bytedance-super-agent-harness/Mon, 23 Mar 2026 12:00:00 +0200https://pinishv.com/articles/deerflow-bytedance-super-agent-harness/37,000 GitHub stars in weeks. #1 on GitHub Trending. ByteDance rebuilt DeerFlow from scratch into a super agent harness with sandboxed execution, sub-agents, persistent memory, and a skills system. It’s not a chatbot framework. It’s closer to what an internal AI platform team would build if they had unlimited runway.https://pinishv.com/articles/zuckerberg-ai-ceo-assistant-obvious-move/Mon, 23 Mar 2026 10:00:00 +0200https://pinishv.com/articles/zuckerberg-ai-ceo-assistant-obvious-move/Mark Zuckerberg is reportedly building an AI agent to help with his CEO duties. My reaction: this is obvious, and frankly late. I’ve been running two AI assistants for a while now, one personal and one for work, and updating them constantly. This shouldn’t be news. It should be default.https://pinishv.com/articles/cursor-automations-ai-stopped-waiting/Mon, 23 Mar 2026 09:00:00 +0200https://pinishv.com/articles/cursor-automations-ai-stopped-waiting/Cursor shipped Automations on March 5. AI agents now trigger from Slack messages, Git pushes, PagerDuty alerts, and timers. No human in the prompt loop. The sequence just changed again.https://pinishv.com/articles/ai-reviewing-ai-code/Mon, 23 Mar 2026 07:00:00 +0200https://pinishv.com/articles/ai-reviewing-ai-code/In the same two weeks, Anthropic launched AI code review, Cursor shipped autonomous security reviews, and GitLab dropped $0.25 agentic reviews. The industry’s answer to ’too much AI code for humans to review’ is ’let AI review it too.’ Where does understanding go?https://pinishv.com/articles/glassworm-invisible-unicode-supply-chain/Sun, 22 Mar 2026 20:00:00 +0200https://pinishv.com/articles/glassworm-invisible-unicode-supply-chain/151 malicious packages in one week. The payload is invisible. Literally invisible. Glassworm uses Unicode characters that don’t render in any editor, terminal, or code review tool. And the cover commits are AI-generated. Here’s how it works and why your current defenses probably miss it.https://pinishv.com/articles/wordpress-ai-agents-publish-web/Sun, 22 Mar 2026 19:00:00 +0200https://pinishv.com/articles/wordpress-ai-agents-publish-web/WordPress.com added MCP write access. AI agents can now draft, edit, and publish posts across 43% of all websites. Meanwhile, YouTube is deleting AI-generated content and demonetizing channels. Two platforms. Opposite directions. Same question: what’s content worth when machines make it free?https://pinishv.com/articles/ai-didnt-replace-software-engineering/Sun, 22 Mar 2026 00:00:00 +0000https://pinishv.com/articles/ai-didnt-replace-software-engineering/The culture shifted. ‘Ship fast with AI’ became the expectation. Anyone who slows down to think looks unproductive. Discipline became a career risk. And that’s how engineering organizations quietly rot from the inside.https://pinishv.com/articles/saas-is-dead-we-just-havent-stopped-paying-for-it/Sat, 21 Mar 2026 00:00:00 +0000https://pinishv.com/articles/saas-is-dead-we-just-havent-stopped-paying-for-it/The bargain that powered SaaS for two decades was simple: rent generic software because custom is too expensive. That bargain is breaking. The cost of building custom software has collapsed, the UI moat is shrinking, and a lot of workflow rent is about to get repriced.https://pinishv.com/articles/google-stitch-design-becomes-infrastructure/Fri, 20 Mar 2026 10:00:00 +0200https://pinishv.com/articles/google-stitch-design-becomes-infrastructure/Figma’s stock dropped 8.8% when Google announced Stitch updates. But the people panicking are asking the wrong question. The question isn’t whether Stitch replaces designers. It’s what happens when a developer can go from idea to interactive prototype in 12 seconds.https://pinishv.com/articles/openclaw-ai-out-of-the-browser/Thu, 19 Mar 2026 14:00:00 +0200https://pinishv.com/articles/openclaw-ai-out-of-the-browser/Everyone keeps comparing OpenClaw to ChatGPT. They’re looking at the wrong layer. OpenClaw isn’t trying to be a better chat UI. It’s trying to move AI out of the browser and into the communication surfaces where you actually live and work.https://pinishv.com/articles/open-webui-ai-interface-infrastructure/Wed, 18 Mar 2026 10:00:00 +0200https://pinishv.com/articles/open-webui-ai-interface-infrastructure/Everyone keeps calling Open WebUI a self-hosted ChatGPT alternative. They’re missing the point. The interesting question isn’t whether it can replace ChatGPT. It’s what happens when the AI interface layer stops being someone else’s product and becomes part of your stack.https://pinishv.com/articles/notebooklm-google-research-workbench/Tue, 17 Mar 2026 10:00:00 +0200https://pinishv.com/articles/notebooklm-google-research-workbench/Everyone compares NotebookLM to ChatGPT. Wrong comparison. ChatGPT starts with a blank chat box. NotebookLM starts with your sources. That difference sounds small. It changes everything about how the tool thinks, what it can do, and where it fails.https://pinishv.com/articles/developer-knowledge-hub-ai-agents-need-context/Sun, 15 Mar 2026 10:00:00 +0200https://pinishv.com/articles/developer-knowledge-hub-ai-agents-need-context/Every AI agent in your org starts every session with zero context. No business rules. No architecture decisions. No conventions. The code they generate looks correct but violates assumptions that live in people’s heads. The solution isn’t better models. It’s a knowledge system.https://pinishv.com/articles/org-charts-for-ai-agents-mapping-your-human-and-ai-workforce/Sat, 13 Dec 2025 15:30:00 +0200https://pinishv.com/articles/org-charts-for-ai-agents-mapping-your-human-and-ai-workforce/AI agents aren’t coming. They’re already here, doing real work, while most organizations are still debating how to use ChatGPT. If you’re not thinking about where they fit in your team structure, you’re already behind.https://pinishv.com/articles/cli-agent-orchestrator-when-one-agent-isnt-enough/Wed, 05 Nov 2025 00:00:00 +0000https://pinishv.com/articles/cli-agent-orchestrator-when-one-agent-isnt-enough/AWS open-sourced CLI Agent Orchestrator, a framework coordinating multiple AI agents for complex developer tasks. It’s hierarchical orchestration for CLI tools, showing where AI tooling is headed when single agents hit their limits.https://pinishv.com/articles/web-bot-auth-crypto-identity-for-bots/Wed, 05 Nov 2025 00:00:00 +0000https://pinishv.com/articles/web-bot-auth-crypto-identity-for-bots/Bots are everywhere, and you can’t tell the real ones from the fakes. Web Bot Authentication uses cryptographic signatures to give automated clients verifiable identities, finally moving past User-Agent headers we’ve been trusting since 1999.https://pinishv.com/articles/ai-wrapper-companies-legitimacy-or-hype/Tue, 04 Nov 2025 00:00:00 +0000https://pinishv.com/articles/ai-wrapper-companies-legitimacy-or-hype/Every company is suddenly an AI company. But when you look under the hood, most are just wrapping OpenAI or Anthropic APIs in a nice UI. Is this legitimate business strategy or temporary hype? And how is this different from companies that built on AWS?https://pinishv.com/articles/amazon-s3-vectors-when-storage-learns-to-think/Fri, 31 Oct 2025 00:00:00 +0000https://pinishv.com/articles/amazon-s3-vectors-when-storage-learns-to-think/AWS just turned S3 into a vector database. It’s not the fastest option out there, but it might be the one that changes how we think about AI infrastructure. Here’s what actually matters.https://pinishv.com/articles/ai-browser-hijacking-how-companies-fight-prompt-injection/Thu, 30 Oct 2025 00:00:00 +0000https://pinishv.com/articles/ai-browser-hijacking-how-companies-fight-prompt-injection/AI browsers that summarize pages and automate tasks are vulnerable to prompt injection—hidden instructions in web content that can hijack the AI. Understanding how this works and what’s being done about it isn’t just useful. It might save you from the next breach.https://pinishv.com/articles/openai-economic-blueprints-what-are-they-doing/Sat, 25 Oct 2025 00:00:00 +0000https://pinishv.com/articles/openai-economic-blueprints-what-are-they-doing/OpenAI just dropped economic blueprints for Japan, South Korea, Australia, the EU, and the US. This isn’t about selling ChatGPT anymore. It’s about reshaping entire economies, and it matters for your career.https://pinishv.com/articles/krakow-offsite-real-connections-real-momentum/Thu, 23 Oct 2025 00:00:00 +0000https://pinishv.com/articles/krakow-offsite-real-connections-real-momentum/https://pinishv.com/articles/securing-intelligence-complete-video-series/Fri, 17 Oct 2025 00:00:00 +0000https://pinishv.com/articles/securing-intelligence-complete-video-series/Don’t feel like reading 15,000 words on AI security? Let NotebookLM read it to you. Sit back, relax, and enjoy the slideshow as we walk through prompt injection attacks, defensive architectures, supply chain risks, and security culture.https://pinishv.com/articles/when-ai-writes-90-percent-of-code/Fri, 17 Oct 2025 00:00:00 +0000https://pinishv.com/articles/when-ai-writes-90-percent-of-code/Anthropic’s CEO says Claude writes 90% of code for most teams. If you think that means developers are obsolete, you’ve missed the point entirely.https://pinishv.com/articles/nvidia-cursor-endorsement/Fri, 17 Oct 2025 00:00:00 +0000https://pinishv.com/articles/nvidia-cursor-endorsement/Jensen Huang name-checked Cursor among six AI startups critical for future work. After a year of using Cursor myself, I understand exactly why Nvidia chose it over everything else.https://pinishv.com/articles/ai-security-culture-problem/Tue, 14 Oct 2025 00:00:00 +0000https://pinishv.com/articles/ai-security-culture-problem/You can implement every technical control and still get breached if your culture doesn’t support security. The final piece of AI security isn’t technology—it’s people, processes, and organizational mindset.https://pinishv.com/articles/securing-the-ai-supply-chain/Mon, 13 Oct 2025 00:00:00 +0000https://pinishv.com/articles/securing-the-ai-supply-chain/We’re building AI systems on top of models, datasets, and tools we don’t control. The supply chain is the attack vector nobody’s defending, and the implications are staggering.https://pinishv.com/articles/building-ai-systems-that-dont-break-under-attack/Sun, 12 Oct 2025 00:00:00 +0000https://pinishv.com/articles/building-ai-systems-that-dont-break-under-attack/Understanding the threats is step one. Building defensive architectures that actually work in production is step two. Here’s what’s working, what’s not, and the trade-offs nobody talks about.https://pinishv.com/articles/prompt-injection-2-0-the-new-frontier-of-ai-attacks/Sat, 11 Oct 2025 00:00:00 +0000https://pinishv.com/articles/prompt-injection-2-0-the-new-frontier-of-ai-attacks/Prompt injection has evolved from toy demos to sophisticated attacks targeting production AI systems. What was once a curiosity is now a genuine security threat that most teams aren’t prepared for.https://pinishv.com/articles/ais-dual-edge-when-to-disrupt-when-to-compound/Wed, 08 Oct 2025 00:00:00 +0000https://pinishv.com/articles/ais-dual-edge-when-to-disrupt-when-to-compound/Your exec team wants ‘AI transformation.’ Your board wants competitive advantage. You need to decide where to deploy your limited engineering capacity. AI has two plays: disrupt or augment. Pick wrong and you waste six months and burn credibility. Here’s how engineering leaders actually make this call.https://pinishv.com/articles/grokipedia-ai-makes-competing-easy/Tue, 07 Oct 2025 00:00:00 +0000https://pinishv.com/articles/grokipedia-ai-makes-competing-easy/Elon Musk announced Grokipedia, an AI-powered encyclopedia built by xAI to compete with Wikipedia. What’s striking isn’t the announcement itself, but how remarkably easy it has become to challenge platforms that took decades to build. Are we seeing the beginning of a complete platform revolution?https://pinishv.com/articles/google-jules-always-on-radar-never-star/Sun, 05 Oct 2025 00:00:00 +0000https://pinishv.com/articles/google-jules-always-on-radar-never-star/Google Jules keeps catching my attention with updates like their new CLI and API, but it never quite becomes the breakthrough tool I’m hoping for. Here’s why it’s interesting but still playing catch-up.https://pinishv.com/articles/gemini-in-chrome-two-weeks-later/Sun, 05 Oct 2025 00:00:00 +0000https://pinishv.com/articles/gemini-in-chrome-two-weeks-later/After two weeks of daily use, Gemini in Chrome has fundamentally changed how I browse the web. Here’s what works, what doesn’t, and why you need a VPN to access it outside North America.https://pinishv.com/articles/ship-faster-without-breaking-things-dora-2025/Sat, 04 Oct 2025 00:00:00 +0000https://pinishv.com/articles/ship-faster-without-breaking-things-dora-2025/AI is making teams faster, but instability persists. The 2025 DORA report reveals which organizational capabilities turn AI into a force multiplier—and which ones let it amplify the mess.https://pinishv.com/articles/build-your-first-ai-agent-this-week/Fri, 03 Oct 2025 00:00:00 +0000https://pinishv.com/articles/build-your-first-ai-agent-this-week/Stop reading about AI agents and build one. Here’s the step-by-step path: picking the right problem, setting up your tools, building a working agent in seven days, and deploying it to your team.https://pinishv.com/articles/build-your-own-ai-agents-for-real-productivity/Thu, 02 Oct 2025 00:00:00 +0000https://pinishv.com/articles/build-your-own-ai-agents-for-real-productivity/Beyond the hype and the demos, what actually works when you build AI agents for real work? Here’s the landscape, the platforms worth using, and what separates success from expensive failure.https://pinishv.com/articles/whats-holding-you-back-from-succeeding-in-the-ai-era/Wed, 01 Oct 2025 00:00:00 +0000https://pinishv.com/articles/whats-holding-you-back-from-succeeding-in-the-ai-era/I’ve watched teams double their output with AI, and I’ve also seen developers stall and managers struggle. The difference isn’t the tools, it’s what gets exposed when AI handles the grunt work.https://pinishv.com/articles/model-context-protocol-connecting-ai-to-your-real-work/Tue, 30 Sep 2025 00:00:00 +0000https://pinishv.com/articles/model-context-protocol-connecting-ai-to-your-real-work/Your AI coding assistant is blind to your company’s actual context. MCP fixes that. Here’s how to connect Claude, ChatGPT, and Cursor to your databases, documentation, and workflows—and why this changes everything about how we build software.https://pinishv.com/articles/agentic-commerce-protocol-when-llms-can-buy-things/Mon, 29 Sep 2025 00:00:00 +0000https://pinishv.com/articles/agentic-commerce-protocol-when-llms-can-buy-things/OpenAI and Stripe just released a new protocol that lets AI agents complete purchases. This is the HTTP moment for AI commerce, and we need to talk about what happens next.https://pinishv.com/articles/the-context-problem-why-switching-between-claude-chatgpt-and-grok-feels-like-groundhog-day/Mon, 29 Sep 2025 00:00:00 +0000https://pinishv.com/articles/the-context-problem-why-switching-between-claude-chatgpt-and-grok-feels-like-groundhog-day/Every time you switch from Claude to ChatGPT, you start from zero. It’s not a bug. It’s architecture. Here’s the real engineering behind AI memory, why context doesn’t transfer, and what it reveals about the future of intelligence.https://pinishv.com/articles/the-magic-behind-ai-ides-how-cursor-windsurf-and-friends-actually-work/Sun, 28 Sep 2025 00:00:00 +0000https://pinishv.com/articles/the-magic-behind-ai-ides-how-cursor-windsurf-and-friends-actually-work/Everyone’s using AI IDEs but few understand what’s happening under the hood. Let’s demystify how these tools work, why they differ, and what’s actually worth paying for.https://pinishv.com/articles/developer-work-did-not-change-the-sequence-did/Sat, 27 Sep 2025 00:00:00 +0000https://pinishv.com/articles/developer-work-did-not-change-the-sequence-did/AI doesn’t make the job different. It changes when parts of the job happen, turning Monday morning from archaeology into editing.https://pinishv.com/articles/github-dual-cli-release-reshaping-development/Fri, 26 Sep 2025 00:00:00 +0000https://pinishv.com/articles/github-dual-cli-release-reshaping-development/GitHub released two different CLI tools for AI in one week. Together, they represent both interactive AI partnership and autonomous development delegation. Here’s why this combination changes everything about building software.https://pinishv.com/articles/so-youve-built-a-rag-now-lets-make-it-not-suck/Tue, 23 Sep 2025 00:00:00 +0000https://pinishv.com/articles/so-youve-built-a-rag-now-lets-make-it-not-suck/Your basic RAG works in the demo. It falls apart with real users. Here’s how to upgrade from a fragile prototype to a production system that actually handles the messy reality of user queries.https://pinishv.com/articles/hiring-developers-in-the-age-of-ai-what-actually-matters-now/Mon, 22 Sep 2025 00:00:00 +0000https://pinishv.com/articles/hiring-developers-in-the-age-of-ai-what-actually-matters-now/LeetCode is dead. With AI writing the code, we need to fundamentally rethink how we identify and hire the developers who will actually thrive in 2025 and beyond.https://pinishv.com/articles/rag-for-developers-a-no-bs-introduction/Sun, 21 Sep 2025 00:00:00 +0000https://pinishv.com/articles/rag-for-developers-a-no-bs-introduction/Developers keep asking me to explain RAG. Here’s the straightforward explanation: it’s the difference between an AI that makes stuff up and one that actually knows your company’s data.https://pinishv.com/articles/the-uneven-reality-of-ai-adoption-what-anthropics-new-report-tells-us/Sat, 20 Sep 2025 00:00:00 +0000https://pinishv.com/articles/the-uneven-reality-of-ai-adoption-what-anthropics-new-report-tells-us/AI adoption has exploded from 20% to 40% of U.S. workers in just two years. But Anthropic’s new Economic Index reveals a troubling pattern: the benefits are concentrating in already-wealthy regions while others fall behind. Here’s what this means for the future of work and inequality.https://pinishv.com/articles/the-context-engine-what-comes-after-weve-solved-code-generation/Fri, 19 Sep 2025 00:00:00 +0000https://pinishv.com/articles/the-context-engine-what-comes-after-weve-solved-code-generation/We’ve largely solved the context problem in AI coding tools. RAG systems ingest our docs, codebase-aware assistants understand our architecture, and operational feedback loops are closing. So what’s next? The real opportunity isn’t building better context engines—it’s leveraging them to fundamentally reshape how we think about software development.https://pinishv.com/articles/im-pro-ai-thats-exactly-why-im-worried-about-our-next-senior-engineers/Thu, 18 Sep 2025 00:00:00 +0000https://pinishv.com/articles/im-pro-ai-thats-exactly-why-im-worried-about-our-next-senior-engineers/A guide for engineering managers on growing junior developers in an AI-heavy world, and for junior developers who want to stand out beyond just being ‘AI operators.’https://pinishv.com/articles/when-ci-cd-speaks-human/Wed, 17 Sep 2025 00:00:00 +0000https://pinishv.com/articles/when-ci-cd-speaks-human/GitHub Next’s Agentic Workflows point to a near-future where we describe CI/CD in plain English and compile it to Actions—auditable, safe, and GitHub-native.https://pinishv.com/articles/from-toys-to-tools-the-missing-layer-developers-actually-need/Tue, 16 Sep 2025 00:00:00 +0000https://pinishv.com/articles/from-toys-to-tools-the-missing-layer-developers-actually-need/AI coding isn’t about clever completions anymore. It’s about stitching work together so results cross the threshold from toy to tool—from interesting demos to outcomes you can trust.https://pinishv.com/about/Mon, 01 Jan 0001 00:00:00 +0000https://pinishv.com/about/https://pinishv.com/work-with-me/Mon, 01 Jan 0001 00:00:00 +0000https://pinishv.com/work-with-me/