Verifiable data platform for distributed compute. Built to make data work wherever compute takes you. Private Beta is live.
- Tweets 905
- Following 16
- Followers 35,508
- Likes 1,444
Moving data has a cost. So does waiting for it.
Transfer and egress land on one bill. Compute waiting for data lands on another.
Price your data path against your compute bill, not just your storage bill.
23
32
106
3,348
When your compute moves and your data doesn’t, you have choices.
Move it. Duplicate it. Or stay where you are.
Shelby is built to give you another option.
Single Address. Global Access.
33
34
169
8,303
This is the part that gets overlooked.
The code might be portable. The data around it often isn't.
That can constrain where the workload runs - or leave compute waiting for the data to catch up.
This is Egress fees, Storage lock-in, Latency to storage, Stateful coupling ( databases, caches, and config baked into a specific environment make workloads sticky even when the app code itself is portable).
14
15
102
5,990
Better compute just became available somewhere else. What stops you moving the workload?
19
26
130
5,711
The cluster section is the one that stands out for us.
Compute capacity only becomes useful when the rest of the system can keep up. That includes the data layer feeding the workload.
As compute becomes more distributed, keeping that data accessible across changing compute environments becomes a bigger part of the infrastructure problem.
We made an interactive explainer of the AI compute stack: from grid power and data centers to chips, racks, clusters, and workloads, plus the public companies at each layer.
Explore the infrastructure behind what @JensenHuang calls the five layer cake:
intelligence.mts.now/compute
24
25
131
10,492
AI compute capacity has been growing at ~3.4× a year.
That’s creating more options than ever for where workloads could run.
Data: @EpochAIResearch
38
33
163
8,376
Compute is increasingly dynamic - where it runs, what’s available, what it costs and what fits the workload.
But changing compute often means another cycle of moving, copying, staging or synchronizing data.
Shelby is built to change that - keeping data accessible as compute changes.
40
35
200
9,051
We’ve been heads down hardening Shelby for Private Beta, and we’re beginning to onboard a select group of customers in stages.
Over the next few weeks, we’ll share more about what testnet clarified, how those learnings shaped the product, and the problem Shelby is built to solve as compute becomes more distributed.
76
45
326
32,764
In March, we opened Early Access to see what builders would do with Shelby's testnet.
You stress-tested it into shape. That was the job, and the job is done. The testnet goes offline on Monday, August 3. What it proved goes live in private production, hardened with customers we onboard directly.
The testnet proved the tech. Private production puts it to work. Then we open it up.
Full rundown in Discord.
1/ AI needs crypto's infrastructure. Crypto needs AI's scale.
Shelby is built for both. Early Access is open. Apply to build today.
67
55
288
38,378
Markets that demand institutional-grade execution. Machines moving faster than any human can. Aptos is the full stack for both.
@DecibelTrade for trading, @shelbyserves for the data layer AI runs on.
$50M+ committed by Aptos Foundation and @AptosLabs: aptosnetwork.com/currents/ma…
338
128
518
37,950
Enterprises don't have a GPU problem. They have a coordination problem: compute in one place, data in another, no efficient path between them.
Building more hardware won't fix that. It compounds the problem.
34
26
162
8,858
$318 billion spent on AI infrastructure in 2025. Less than 2.5% went to storage. (IDC, 2026)
The industry scaled compute.
Nobody fixed how data gets to it.
38
40
194
12,840
Last night, one fibre cut in eastern North America sent latency and timeouts as far as Europe, taking major platforms down with it. A single physical fault, a continent-wide blast radius.
Data shouldn't go dark because a cable did. The bytes were fine the whole time.
29
27
163
10,061
Regional storage is as 2021 as a return-to-office mandate. From one place, you can be everywhere.
So why does your data still have to be regional?
Shelby. Write once, read from anywhere on Earth.
45
43
272
760,521
