About Minute.ly
A small team working on AI and video.
Minute.ly is a small team working where artificial intelligence meets video.
We started publishing here because most writing about AI video is not very good. It comes either from vendors describing their own products, or from people who have never had to explain to a publisher why the numbers did not move. Both produce the same thing: confident claims, no stated limits, and nothing a working team can act on.
We wanted somewhere that said what genuinely works, what does not, and — the part almost everyone skips — where the boundary between them falls.
What we write about
- AI & Technology — how machine learning is actually applied to video: content analysis, frame and preview selection, personalisation, real-time packaging. And where each approach stops working.
- Video Strategy — how teams plan, package and measure video. Which levers move views, watch time and revenue, and which only look like they do.
- Tools & Workflow — practical guides to the tools around video and audio, and a clear view of which tasks are worth handing to a machine.
What we think
A few positions run through most of what we publish.
- Presentation is badly underrated. Most video that goes unwatched is not weak work — it is well-made work that was presented carelessly, because presenting it properly does not scale by hand.
- Automation belongs on mechanical judgements. The repeated, low-stakes decision made ten thousand times is exactly what a model should handle. Deciding what to cover, or what a story means, is not that kind of decision.
- Most video metrics are worse than they look. Counting video starts rewards autoplay and interstitials while the thing you actually depend on — a reader willing to watch video on your site — quietly degrades.
- Stated limits make advice usable. Every technique worth using has a boundary. Naming it is what separates a recommendation from a sales pitch.
How we write
- Specific. A claim that cannot be checked is not worth publishing.
- Bounded. We say where a technique stops working, not only where it starts.
- Our own. We do not borrow other people's results to make a point.
What we're building
Our product work is underway. We will write about it here when there is something real to describe, rather than in advance of it.
Get in touch
If you work with video at any scale, we would like to hear from you — and if something here is wrong, we would especially like to hear about that. Contact us.