AI Agent

Every few months, AI produces a new gold rush narrative. This time it is voice agents. The pitch is seductive because it sounds so clean. Real-time voice. Low latency. Natural conversation. Multilingual support. A few lines of Python. Connect it to a calendar or booking system. Replace a receptionist, scheduler, or tier-1 support role. Charge a monthly subscription. Print money before the market catches up. At the surface level, this sounds plausible. In fact, part of it is true. The underlying models really have improved. Voice interaction is becoming fast enough, natural enough, and cheap enough that many routine call flows are now technically possible. That is not hype. It is real progress. ...

Today I made a structural shift in my OpenClaw setup with the help of Codex-5.3. For those who know my current architecture, I’ve been running Codex CLI in the backend to manage the OpenClaw instance on my Pi. Over time we’ve built it into a complete knowledge holder for everything related to OpenClaw, and that accumulated context is what made this change possible. I had a single strong agent running KOL and content execution, and it worked well. The strain started when I pushed it beyond that scope. Content strategy, software architecture, and private daily support were all forced to share the same brain. ...

Gartner says 40% of enterprise apps will embed AI agents by end of 2026. But here’s the thing nobody’s talking about: agents aren’t just better tools—they’re a new category that breaks how we organize work. The MIT/BCG 2025 research reveals four tensions every leader will face, and most aren’t ready.

Natural language interfaces for data are no longer experimental—they’re becoming essential enterprise tools. But building a reliable, production-grade data agent requires moving beyond simple prompt engineering. This post shares architectural lessons learned from building an enterprise natural language to SQL (NL2SQL) agent. The Problem with Pipeline Architectures Most NL2SQL systems start with a pipeline approach: User Question → Intent Detection → RAG Retrieval → SQL Generation → Validation → Response This works for demos but creates problems at scale: ...

A response to the observation that “coding agents are general agents” and that program synthesis will outperform hand-crafted vertical-specific agents. The Post The Thesis The tweet makes a compelling point: coding agents that write and execute code represent a form of scalable search. Rather than encoding years of expert knowledge into prompts and rules, let the agent explore the solution space through code generation and execution. This is “bitter lesson adjacent”—general methods that leverage computation outperform hand-crafted approaches. ...

The process we inherited I have been rethinking how we clarify requirements in software teams, because the way most of us still work feels optimized for a world that no longer exists. Traditionally, we treated requirements as something that had to be settled before implementation began. We wrote documents, scheduled planning sessions, debated edge cases in meetings, aligned on tickets and estimates, and only after everyone felt comfortable did we finally start building. That process made sense when writing software was expensive and slow. If a feature took weeks to implement, investing time up front to reduce mistakes was rational. ...

A retrospective on upgrading Clawdbot from a reactive task-doer to a proactive, system-thinking partner.

The Wall Street Journal recently reported on the rapid rise of Claude Code and how quickly it has spread beyond the usual circle of AI enthusiasts. That article should cause many CEOs and technology leaders to pause and reconsider the AI strategies they have already committed to. For the last couple of years, most enterprises have followed a very familiar playbook. They selected an approved corporate platform, standardized on a vendor stack such as Gemini Enterprise or Microsoft Copilot, launched formal transformation programs, and brought in outside partners to implement carefully defined use cases. This approach made sense because it mirrored how organizations adopted cloud computing, analytics, and ERP systems. It was orderly, governable, and predictable. ...