Fitting Translated Speech Back Into the Original Timeline: The Duration Problem in AI Dubbing
Everyone who builds an AI dubbing pipeline hits the same wall in week two. The ASR works, the...
Tutorials
Silent failures in agentic systems: when retry and fallback fake an OK
No alert isn’t no problem. Silent retry and fallback return “OK” while quality tanks. Datadog: 1/3 of LLM errors are hidden rate-limits.
XP Is More Important Now Than It Was Before AI
AI coding tools make Extreme Programming practices more important, not less. When the machine writes the code, human judgment about what to build becomes the scarce resource.
Claude Code was mailing your email address to every website it visited: the User-Agent leak, the 5-month system-prompt issue, and the 5-point check
Claude Code put your account email in the User-Agent header of its web requests, so every server log on the agent's path harvested a scam-ready lead, and the same email ships in th...
-U0, -U3 or --function-context: Count the Tokens Your Review Agent Reads per Diff
Same commit, seven diff views, seven token bills. A script that counts what your review agent reads at -U0, -U3, -U10 and -W, and what each one hides.
Real-World TTFT & Throughput Benchmarks across Major LLM Hosted Endpoi: Deep Dive & Engineering Breakdown
Decoding LLM Inference Economics: Real-World TTFT & Throughput Benchmarks Across Hosted...
Ollama vs vLLM: Death Battle 🥊 - 2026
Ten rounds. One verdict. No survivors. You've got a GPU. You've got a model. You need something...
The model card said 0.5B. The checkpoint stores 494M. Both are right.
Qwen2.5-0.5B's config declares its vocabulary-by-hidden matrix twice: once as the embedding table,...
I Built a Trading Bot with AI. It Lost 2.2%.
I gave an AI coding agent one job: build me a trading bot. Not a toy script. A Python project that...
AI agents are your new customers. I built a checker to see which stores they can actually read
A shopper asks ChatGPT for "a waterproof hiking jacket under $150 that ships to Dubai". The answer...
When an AI Agent Says 'Done,' What Exactly Has Finished?
"Done." It's probably the most satisfying word in a coding-agent session. It can also conceal three...
The instrumentation pattern: an agent you cannot see is an agent you cannot operate
Most agents are instrumented for the failures that announce themselves and blind to the ones that do not. Here is the pattern: the three streams...
Don't Scale Your LLM Routing on Request Volume — Scale It on Task Success Rate
Your request volume doubled this month. So you doubled your GPU spend. That is the most common...
MCP Went Stateless This Year. Your 'Supports MCP' Checkbox Is Now a Lie.
"We support MCP" used to mean something specific. This year it became ambiguous, because there are...
I missed /buddy, so I made a pixel pet for Claude Code
In April, Claude Code had /buddy for about a week: a tiny creature in your terminal that watched you...
Celery Containers and Asynchronous ML Microservices : Production Infrastructure Architecture
Building resilient, high-concurrency machine learning (ML) and large language model (LLM) platforms...
Building an AI for hard conversations: five agents, one safety regex, and a supervisor who graded us blind
Context: Dicamus is an AI self-help service, not therapy and not a crisis line. If you are in danger...
I got tired of pasting API keys into every coding agent, so I built a local router
You open Cursor and paste an OpenAI key. Then Claude Code wants Anthropic. OpenCode wants something...
What is your role in your agentic setup?
A few weeks ago I wrote about how I stopped coding and started orchestrating. Ten terminal tabs,...
How I stopped Claude from bloating my codebase
Coding agents are good at adding code and bad at deciding not to. That was my problem in one...
I Tracked Every AI Coding Mistake for a Week — the 5 Patterns That Keep Coming Back
I spent a week building a small side project with an AI coding assistant doing most of the typing....
My AI agent says "all fixed." 3 times out of 4, it never measured the before.
GitHub: https://github.com/ESeungJun/fix-proposal "All fixed." I've heard that from my AI agent,...
My Support-FAQ Bot Made Up 22 Answers It Didn't Know. Grounding It in n8n Dropped That to 0
A support bot that says "I don't know" is annoying. A support bot that confidently says "No, we don't...
Deterministic delivery is an architecture decision, not a model setting
1. The argument is being filed under the wrong heading Most debates about agent...