What this is and who built it
The site
Inference Engineering is a technical reference for people building with or deploying large language models. It covers the concepts that sit between the research and the production system — the parts that don't always make it into documentation but matter a lot when things go wrong or cost too much.
Each guide tries to explain why something works, not just what it is. No ads, no sponsored content.
The author
I'm EMF — based in Australia. My background is in systems engineering, R&D, and enterprise technology. I've spent a lot of time at the point where an idea that works in a lab has to actually run somewhere, and that gap is usually where the interesting problems are.
This site is how I organise what I've learned and keep thinking through it. If something here is wrong, I'd rather know.
Get in touch
If something here is wrong, incomplete, or you just want to talk about any of it — use the form below.
Disclaimer
// Important — please read
Everything published on this site represents my own views, formed independently through research, personal experience, and publicly available information. Nothing here reflects the views, positions, or opinions of any current or past employer.
No proprietary information, internal systems, confidential data, or intellectual property belonging to any employer or client has been used in producing this content. All content is derived from publicly available sources, published research, and general domain knowledge.
This site is not affiliated with, endorsed by, or connected to any organisation I am or have been employed by. The content is produced entirely in a personal capacity, outside of any professional obligations.
Nothing on this site constitutes professional advice of any kind — technical, legal, financial, or otherwise. The field moves fast; information may be incomplete or out of date. Use your own judgement, and verify anything that matters before relying on it.