-
Continue reading →: Reverse Engineering BLE Media Wall Lights for HomeKitTL;DR: I integrated the proprietary BLE media wall lights that came with our new media wall into Apple Home using a Raspberry Pi, Node RED, Python with Bleak, and a Nordic nRF52840 BLE sniffer. The big lesson was simple: packet capture beats protocol guessing when a device has a real…
-
Continue reading →: Building Open Duck Mini V2: From Software Project to Physical RobotTL;DR: I have started building Open Duck Mini V2, an open source humanoid robot, and I am documenting the whole process as an engineering journal. What began as “this looks fun” is quickly becoming a hands on lesson in 3D printing, mechanics, electronics, wiring, calibration, and software. I spend a…
-
Continue reading →: Smart Home Troubleshooting Why the Fix Was a Queue Not a RetryTL;DR: The fix for my flaky Bluetooth blinds was not another retry or a random percentage tweak. It was a proper command queue that debounced noisy updates, serialized BLE access, and treated the Raspberry Pi Bluetooth adapter as a shared resource. A couple of weeks ago, I thought I had…
-
Continue reading →: How to Fix Concurrent BLE Commands on Raspberry PiTL;DR: The blinds were not failing because the Raspberry Pi was simply “too busy.” The real problem was concurrent BLE commands colliding at the Bluetooth stack, leaving BlueZ in an unreliable state until a reboot cleared it. I thought I was chasing a general performance problem. The Raspberry Pi had…
-
Continue reading →: 3 AI Agent Skills That Transformed My Personal ProjectsTL;DR: The three AI agent skills that have had the biggest impact on my personal projects are grill-me, caveman and thermo-nuclear-code-quality-review. Each improves a different part of the development process: thinking clearly, building quickly and reviewing code more rigorously. Over the past few months, I have been experimenting with reusable…
-
Continue reading →: Streamline Workflows: A Guide to Managing Agentic Workflows with Git WorktreeTL;DR: git worktree enables multiple working directories from a single Git repository, facilitating parallel development and sandboxing for agentic workflows, especially beneficial for coding agents. In the fast-paced world of software development, managing multiple tasks simultaneously can be a challenge, especially when working with coding agents. Enter git worktree, a…
-
Continue reading →: Book Review: How the Mighty Fall by Jim CollinsTL;DR: How the Mighty Fall by Jim Collins reveals that organizational decline often happens gradually and invisibly, following a predictable pattern that can be detected and reversed if caught early. The book outlines five stages of decline, highlighting that success can lead to arrogance, undisciplined growth, denial, panic, and eventual…
-
Continue reading →: AI Agents in Your Pocket: Testing Codex on MobileTL;DR: Codex on mobile transforms your phone into a powerful remote control for coding, enhancing productivity by allowing you to manage your existing development environment from anywhere, rather than turning your phone into a standalone coding device. Coding on a mobile device has always seemed appealing but often falls short…
-
Continue reading →: Optimising OpenAI SDK for Dynamic LLM Models in Production AppsTL;DR: To effectively use OpenAI SDK in production apps, treat LLM models as evolving dependencies. Manage model changes with configuration files, evaluate upgrades with tests, and encapsulate API calls in a service layer to minimize disruptions. In the fast-paced world of AI, deploying language models in production apps requires a…
-
Continue reading →: Exploring dry4clj, Uncle Bob’s Clojure tool for detecting repeated code structures and thinking about the DRY principle beyond simple copy-paste duplication.TL;DR: dry4clj, a tool by Uncle Bob, helps detect structurally similar Clojure code, highlighting the importance of the DRY principle beyond mere copy-paste duplication. This becomes crucial as AI-generated code proliferates, emphasizing the need for maintainable and efficient codebases. In the world of software development, the DRY (Don’t Repeat Yourself)…
-
Continue reading →: AI Is Not Replacing Software Engineers – It Is Creating Bottleneck GeneratorsTL;DR: AI is not replacing software engineers; it’s transforming them into bottleneck generators. While AI can accelerate code generation, it doesn’t automatically streamline the entire software development process. The future lies in human and AI collaboration, where AI speeds up tasks and humans provide critical judgment and oversight. In the…
-
Continue reading →: How “Caveman” helps AI coding tools waste fewer tokensTL;DR: Caveman optimizes AI coding tools by enforcing a concise communication style, reducing token waste without sacrificing technical clarity. This approach enhances efficiency in tools like GitHub Copilot by streamlining responses. In the world of AI coding tools, efficiency is key. As developers increasingly rely on these tools to streamline…
-
Continue reading →: Java Mutation Testing: Ensuring Your Tests Detect Bugs EffectivelyTL;DR: Mutation testing in Java enhances the effectiveness of your test suite by introducing small code changes to identify weaknesses in your tests. Tools like PIT automate this process, providing a more meaningful assessment than traditional code coverage metrics. Introduction: In the world of software development, ensuring that your tests…
-
Continue reading →: Unlocking Regime Shifts: A Guide to Rolling StatisticsTL;DR: Understanding investment performance requires more than relying on a single long-term average. By using rolling statistics, investors can observe how return and risk evolve over time, revealing how different market regimes impact performance. Investors often look at historical averages to gauge future performance, but this can be misleading. Markets…
-
Continue reading →: Returns with Reduced Volatility: Smoother Charts, Same Risk ExplainedTL;DR: Resampling financial data to lower frequencies can create the illusion of reduced risk by smoothing out volatility, but the underlying economic risk remains unchanged. It’s crucial to use consistent data granularity when comparing investments to avoid misleading conclusions. Understanding how data frequency affects perceived risk is essential for investors.…
-
Continue reading →: Investment Insights with Log Returns for Risk and Return AnalysisTL;DR: Logarithmic returns provide a consistent method for annualising risk and return, allowing for accurate scaling across time horizons. This approach ensures theoretical precision in financial analysis, distinguishing itself from common industry practices. Understanding financial performance requires more than just looking at raw numbers. By using logarithmic returns, investors can…
-
Continue reading →: Evidence of Fat Tails in Stock Returns: Normality Testing InsightsTL;DR: Financial returns, such as Microsoft’s daily log returns, often deviate from a normal distribution, exhibiting fat tails. This implies that extreme market events occur more frequently than predicted by mean-variance analysis, necessitating additional risk measures. Understanding the behavior of stock returns is crucial for investors and analysts. While traditional…
-
Continue reading →: Risk-Return Analysis with Log Returns: Key Concepts and ChallengesTL;DR: Log returns offer a statistically sound method for comparing asset performance across different markets, providing a clearer picture of risk-return dynamics. However, mean-variance analysis, while useful, falls short due to its assumptions about normal distribution, overlooking the realities of market behavior. Understanding the nuances of risk-return analysis is crucial…
-
Continue reading →: My Honest Review of OpenAI Codex IDE: Is It Worth It?TL;DR OpenAI’s Codex IDE offers a focused, seamless way to integrate AI into coding workflows, especially for existing projects, though it requires careful oversight for quality and security—making it a strong alternative to tools like Cursor or Gemini for Mac users. Introduction Switching AI tools can feel like chasing the…
-
Continue reading →: 2025 Year Review – Why Focus Matters: Key Lessons from a Year of Big MovesTL;DR In a year of big personal changes, I learned that focus is finite, and chasing too many goals leads to burnout rather than breakthroughs—prioritizing high-impact work over busywork is the key to real progress. Introduction Ever feel like you’re juggling a dozen balls, only to watch them all crash…
Hello,
I’m Sean
Welcome to the Scalable Human blog. I’m a software engineer who likes building things, from software and AI tools to robots, Raspberry Pi projects, smart-home systems and energy monitoring. I learn in public, experiment with technology and share what I discover along the way.
Let’s connect
Join the mailing list
Stay updated with latest posts.
Scalable Human Blog 