Category: Uncategorized

A Free AI Knowledge Base for SARS Tax Rules South African tax rules are dense, frequently updated, and spread across PDFs that are not designed for quick lookup. SARS publishes guides, interpretation notes, and practice notes as separate documents. Finding the right answer to a specific VAT or provisional tax question means knowing which document […]

A Private, Fast AI Expert on AWS Bedrock When a client needed an internal Q&A assistant over proprietary documents — technical specs, compliance policies, historical project notes — the requirements were clear: no data leaves the organisation’s AWS environment, responses must be fast enough for interactive use, and the system must be maintainable without a […]

The Sim Studio Integration Failure: a Post-Mortem Not every integration story ends with a working system. This one ends with a decision to remove the integration entirely — and that decision was the right call. The post-mortem is worth writing because knowing when not to add complexity is a more valuable skill than knowing how […]

Fixing Postiz on Docker (Setup + the X Media-Upload Bug) Postiz is a self-hosted social media scheduler with a clean UI and decent multi-platform support. Getting it running on Docker is straightforward once you know the environment variable gotchas. Getting it to successfully upload media to X (formerly Twitter) requires a second round of debugging […]

A Dockerized Remotion Render Pipeline Remotion is a React-based video renderer that produces MP4s from JSX. It works beautifully on a developer’s machine and falls apart the moment you try to run it in CI or hand it off to a server. The problem is Chromium: Remotion uses headless Chrome for rendering and Chrome has […]

Downloading ComfyUI Models Without the Pain Hugging Face has a model download UX designed for interactive notebook sessions, not production pipelines. If you’ve ever watched a 12 GB checkpoint stall at 94% with a 403 error, waited for the browser download bar to reach 8 GB before realising the connection dropped, or tried to resume […]

RTX 3090 ComfyUI Tuning: xformers + sage-attention Default PyTorch attention is safe and predictable. It is not fast. On the RTX 3090 with SDXL, switching from vanilla attention to xformers cut per-image generation time by roughly 25%. Adding sage-attention on top trimmed another 10–15% on longer diffusion runs. Neither change requires touching a single ComfyUI […]

ComfyUI on an RTX 3090: a Local-First Setup Running image generation locally isn’t about avoiding API costs — it’s about having a workflow that doesn’t change under you. Cloud services deprecate models, change pricing, and throttle batch jobs. A local ComfyUI setup on an RTX 3090 stays the same until you decide to change it. […]

Running Claude Code with GLM-4.6 (Ubuntu + Codespaces) Claude Code doesn’t have to run against the Anthropic API. It supports any OpenAI-compatible endpoint — which means you can point it at a locally hosted model. GLM-4.6 from Zhipu AI is a capable open-weight model that speaks the OpenAI chat format, and getting it working with […]

How I Built an AI Dev Framework as a Coding Student I was six months into learning to code when I decided the problem wasn’t what I was building — it was my environment. Every tutorial assumed a clean machine with one project. I had four half-finished projects, conflicting Python versions, and a VSCode install […]