Daniel Rosehill Blog
Exploring technology, AI, business insights, and personal reflections from a curious mind
24 Cheshvan 5787
Improving a personal ASR model by using it every day
How to improve a personal speech-to-text model by using it daily: an agent reviews your transcript history, you confirm the errors, and small fine-tunes follow. A playbook from Parakeet and Hebrish, with a PDF and a Claude Code skill.
23 Cheshvan 5787
Claude Code plugins for non-coding work: homelab, data, media and admin
A tour of my public Claude Code plugins for non-coding work: Proxmox and Synology admin, data analysis and cleaning, image and audio production, budgeting, careers, and idea evaluation, with install steps.
23 Cheshvan 5787
Gerber Prime 6000 in Home Assistant: local control, no cloud
How to get a Gerber Prime 6000 (PSC05) relay and shutter controller into Home Assistant locally: setup Wi-Fi, its HTTP API, moving it onto your LAN, and an MQTT bridge with auto-discovery. With code.
23 Cheshvan 5787
What my cheap AliExpress camera sends to the cloud (and what a Reolink sends instead)
Packet captures of a Xiongmai-based AliExpress camera and a Reolink E1 Pro: hard-coded DNS, four clouds, a plaintext cloud ID, and live-view audio relayed unencrypted. Schematics, scores and how I locked it down.
23 Cheshvan 5787
My curated AI indexes on GitHub: what each list covers
Fifteen curated AI lists and awesome-lists I keep on GitHub: agents and harnesses, coding tools, evals, LLM councils, simulations, voice and Whisper. What each covers, how big it is, when it was last updated and which are snapshots.
Popular tags
All tags →
23 Cheshvan 5787
Can image models spell the labels? A first look at my Tech Graphics Benchmark
Limited evaluation: 95 images from five cheap image models, 12 annotated so far. In this run, 10 of 11 had every label spelled exactly; the problems were composition, stretching and style. Not a verdict on any model.
23 Cheshvan 5787
Datasets for fine-tuning speech recognition on your own voice
A guide to my public Hugging Face datasets of read-aloud sentences for fine-tuning speech recognition: tech, Linux, smart home, AI agents, homophones, Hebrew code-switching and voice notes. What each holds and which to use.
23 Cheshvan 5787
Transcribe a WhatsApp chat export, voice messages included
How to read and transcribe a WhatsApp chat export: WhatsApp Export Unpacker parses the ZIP, transcribes voice messages with Whisper, and builds a text file, PDF and WhatsApp-style HTML viewer, with optional anonymisation of participants.
23 Cheshvan 5787
My AI prompt libraries, and which one to use
A guide to my open-source system prompt library and its siblings: prompts for NotebookLM, Open WebUI, Roo Code and ChatGPT, tools for writing prompts, and which of the twelve repositories and Hugging Face datasets to use for what.
23 Cheshvan 5787
Obsidian AI plugins, sorted: a categorised list of 86 LLM plugins
A categorised list of 86 Obsidian plugins that bring AI and LLMs into a vault, from local Ollama setups to chat-with-notes RAG and voice input, sorted by GitHub stars. Auto-organised from GitHub metadata, not hand-reviewed.