After 1 Year of Use: The Gemini API in 2026
After one year of use in my production environment, the Gemini API has proven itself to be a mixed bag—useful for small projects but a headache for scaling larger systems. If you’re keen to know what makes this API tick, read on and brace yourself
After 6 months with LangGraph in production: it’s good for rapid prototyping, painful for scaling to enterprise.
I’m no stranger to adapting to new platforms as a developer, but let me tell you, LangGraph has been quite the ride for the last half a year. Launched by the GitHub user langchain-ai, this tool aims
Hey everyone, Kai Nakamura here from clawdev.net, your usual spot for all things AI development. Today, I want to talk about something that’s been on my mind a lot lately, something that many of us interact with daily, but perhaps don’t give enough thought to when it comes to our own contributions: the often-overlooked art
How My Journey in OpenClaw Began
I’ll never forget the first pull request I made to OpenClaw; the nerves felt like a million insects crawling under my skin. I had spent hours perfecting my code, convincing myself I had seen every edge case. Then boom! A single typosaurus reared its ugly head – I’d written
Most guides about building multi-tenant setups in OpenClaw are wrong. Seriously, you might end up pulling your hair out following some of them. When I first tried deploying a multi-tenant environment, I got stuck for hours trying to juggle configurations like a circus clown. Imagine this: you’re knee-deep in YAML files, and suddenly your whole
Hey everyone, Kai Nakamura here from clawdev.net, and today I want to talk about something that’s been buzzing in my Slack channels and GitHub feeds for weeks: the quiet revolution happening in open-source AI development. Not the big, splashy foundation model releases, but the nitty-gritty, the tools, the infrastructure, the stuff that makes building with
7 Performance Profiling Mistakes That Cost Real Money
I’ve seen 15 applications slow down significantly in the last quarter, and guess what? All of them made the same 7 performance profiling mistakes. These mistakes don’t just waste developers’ time; they can cost companies a fortune in lost productivity, infrastructure fees, and customer churn. Understanding what
TGI vs llama.cpp: A Showdown for Small Teams
Start with this: TGI from Hugging Face has 10,811 GitHub stars, while llama.cpp lags behind. But let’s be real—stars don’t mean squat if the tool doesn’t get the job done. In a world where small teams need efficiency but lack the luxury of extensive resources, the choice
Listen Up: Picking Up Where I Left Off
I remember my first time contributing to the OpenClaw project. It was exhilarating and terrifying all at once. I had zero clue about what lay behind those hundreds of files. But once I cracked open the code and started fiddlin’ with plugins, it felt like discovering a
Hey everyone, Kai Nakamura here from clawdev.net! It’s been a minute since I dove deep into the nitty-gritty of what makes our AI dev world tick, and today, I’ve got something that’s been on my mind for a while. We talk a lot about building models, training data, and optimizing algorithms, but what about the