co-founder, cto @ north.cloud +$800M/yr of cloud spend under management
USA
- Tweets 1,088
- Following 190
- Followers 118
- Likes 442
Açoine retweeted
OX Alpha is hosted in America!! 🇺🇸
I sent 49 identical one-token requests from 7 Cloudflare regions, rotated the order to control for load and matched every response to OpenRouter’s per-generation server timing.
This isolated the geographic network floor:
ATL: 57 ms
DFW: 76 ms
Europe: 125–132 ms
Asia: 202–340 ms
Australia: 279 ms
A robust 6-of-7 fit converged at 43°N, 73°W, with the plausible band covering the US East Coast into southeastern Canada.
Best interpretation: OX Alpha’s serving path is anchored in eastern North America, most likely US East.
The only potential remaining caveat: this could be @OpenRouter backend handoff rather than the physical GPU rack.
131
65
2,706
299,136
I disagree.
First your definition of a cost center doesn’t make sense here. A company saving money for another company can be cost cutting for either a cost center or a profit center. How it impacts a profit center vs cost center might differ, but if you roll that up to COGs for a company it’s all the same.
I digress. Your initial point was that cost cutting cannot scale like revenue generation. The same issue that impacts your max savings potential per customer also impacts your max revenue potential per customer.
One is called TAM for a company the other is spend. They are independent of each other but related.
If the TAM increases both the cost savings and revenue generation potential increases. If the TAM decreases the pressure for savings is accentuated not the opposite.
I don’t think one is better than the other but when your revenue drops or stagnates the first thing companies try to do is to cut costs not try to do the thing they’re failing at.
Hot take: If you are building a business that helps companies save $$, your upside will always be capped and eventually you will be commoditized.
Nobody will pay you more than the cost you save them.
But if you help a company make more money, there is no ceiling to what they will pay you.
If you are serious about building a massive business, find a way to sit on the revenue-generating side of the business, not the cost center.
1
109
Thanks for feature !
☁️ Cloud costs are rising, but north.cloud is helping companies cut cloud & AI spend by 30%+, often within the 1st day.
CTO & co-founder @AcoineY shares more in Founders Everywhere
🔗 ideas.everywhere.vc/p/north-…
Also, they're hiring: north.cloud/careers
ALT A company profile card for North.cloud with a magenta header displaying the mirrored “EVERYWHERE” Ventures logo. On the left, the North.cloud logo appears above the label “Future of Work” and the description: “The financial operating system for your cloud, AI, and data infrastructure.” The website “NORTH.CLOUD” is listed at the bottom left. On the right, there is a circular black‑and‑white portrait labeled “Yassine Açoine, CTO, CO‑FOUNDER.”
4
75
I want to write a piece about about the importance of writing and reading as a technical person in the age of AI.
.I think the best engineers are not technical but amazing at writing and reading. Same advice I share to all investors that ask me what their kids should focus on.
AI interactions are text based. In order to route to the best outcome, the most efficiently, you have to be precise in your language and be able to read complex answers returned by AI.
41
Something does not make sense to me. If companies are increasing their AI spend exponentially but do not see ROI / real benefits, then this is the only set of possibilities:
1. Increasing feature velocity does not correlate to increased revenue.
2. Giving engineers more speed to deliver does not make the product better. It is the product improvements that make the product better. The vision/taste around how a feature is used and consumed.
3. Engineering or product efforts are directed toward the wrong things.
4. There is a true delay between engineering improvements and the GTM motion, meaning GTM needs to adopt AI more to proportionally deliver the message.
5. Engineers are replacing the grunt work of coding with AI and are not using that net new time with creative approaches.
6. Speed of delivery does not mean better engineering. A shit idea delivered 10x faster is still a shit idea.
7. This is most likely what is happening, in my opinion. The incentive was set incorrectly, where executives correlate LLM spend to better engineering, which then incentivizes engineers to blindly prompt instead of being slower and more deliberate about AI usage.
8. The limits of AI's benefits are bound by the user's proficiency and intelligence and we are seeing our own limits.
9. Enterprise software is extremely complex, and because of the memory and context limitations that AI has, the benefits it can deliver are greatly reduced. This is why it is easier to deliver an MVP through one-shot, but much harder to deliver a feature improvement against an existing codebase. LLMs are context hungry at a certain point; codebases are massive, not intuitive, and require a lot of guidance.
Some of these points are contradictory on purpose.
1
4
55
Açoine retweeted
Our engineers are working around the clock to cut your cloud costs. Just another productive sprint.
#finops #aws #gcp #engineeringteam
1
68