https://yosgi.github.io/en/tags/ai-systems/Recent content in AI Systems on YosgiHugoenMon, 29 Jun 2026 09:20:51 +0000https://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-3/Sun, 21 Jun 2026 10:00:00 +1200https://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-3/Making scan results land in the right placehttps://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-2/Sat, 20 Jun 2026 19:11:11 +1200https://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-2/How we split 2D detection and 3D verification into two stages to reduce false positives in digital twin asset annotation, and how that revealed the next recall bottleneckhttps://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-1/Fri, 19 Jun 2026 09:46:11 +1200https://yosgi.github.io/en/post/using-ai-agents-for-asset-recognition-and-annotation-in-3d-scenes-part-1/How we turned free-form 3D exploration into measurable coverage scanning and improved valve inventory recall in a digital twin factory from around 40% to around 90%https://yosgi.github.io/en/post/from-extracting-drawing-text-to-placing-2d-annotations-in-a-3d-scene/Fri, 12 Jun 2026 11:28:21 +1200https://yosgi.github.io/en/post/from-extracting-drawing-text-to-placing-2d-annotations-in-a-3d-scene/A practical engineering recap of how a simple drawing text extraction task turned into a 2D-to-3D mapping workflow for digital twins, combining SVG parsing, vision models, geometry rules, registration math, and agent tool design.https://yosgi.github.io/en/post/why-tool-and-code-fail-differently-in-agent-systems2/Mon, 25 May 2026 14:50:20 +1200https://yosgi.github.io/en/post/why-tool-and-code-fail-differently-in-agent-systems2/Tool and Code do not just differ in expressiveness; they also fail at different times. This article explains why multi-round Agent systems amplify that difference.https://yosgi.github.io/en/post/tool-for-control-code-for-analysis3/Mon, 25 May 2026 14:50:20 +1200https://yosgi.github.io/en/post/tool-for-control-code-for-analysis3/In large-scale Agent systems, pure Tool-based approaches tend to collapse under context pressure, while pure Code-based approaches introduce excessive latency. Based on real-world experiments, this article introduces a dual-path execution model that uses a “Context Off-Ramp” to switch between Tool and Code execution.https://yosgi.github.io/en/post/an-mcp-as-code-refactor-and-why-it-did-not-work-the-way-i-expected1/Mon, 25 May 2026 14:50:20 +1200https://yosgi.github.io/en/post/an-mcp-as-code-refactor-and-why-it-did-not-work-the-way-i-expected1/A real-world MCP-as-Code refactor on where Code helps, where Tool still fits better, and why Agent runtime needs both.https://yosgi.github.io/en/post/ai-won-t-replace-software-engineers-it-will-replace-the-parts-of-coding-we-never-loved/Thu, 21 May 2026 23:28:56 +1200https://yosgi.github.io/en/post/ai-won-t-replace-software-engineers-it-will-replace-the-parts-of-coding-we-never-loved/AI will not replace software engineers as a whole. It will remove much of the repetitive, mechanical work and give engineers more leverage to focus on judgment, systems, users, and building better products.https://yosgi.github.io/en/post/your-agent-is-not-inconsistent-because-it-is-dumb-it-just-has-too-many-tool-paths/Thu, 21 May 2026 18:20:04 +1200https://yosgi.github.io/en/post/your-agent-is-not-inconsistent-because-it-is-dumb-it-just-has-too-many-tool-paths/Multi-tool agents become unreliable when the same task takes different tool paths each time. This article explains why Skills should preserve validated execution paths, not one-off results, and why those Skills need to evolve over time.https://yosgi.github.io/en/post/early-experiences-with-building-mcp-tools/Thu, 29 Jan 2026 17:45:36 +1300https://yosgi.github.io/en/post/early-experiences-with-building-mcp-tools/How Cut 6,000 Tokens Down to 500