Serverless wallet?

I’ve done various iterations of this concept over the years.
First think about transaction latency over multiple regions, ie CA to NY with the usual deployment in the middle of the country.
That can be a problem with data in flight.
Serverless is less about having full traditional access to a VM, but rather an abstract like a key value store that you can extend.
In this case it’s DynamoDB, but there are many data products out there that have this capability by being presented to your users by edge data centers closest to them.

You’ll note most of the transactions are being handled by Lambda functions rather than discrete compute.
This allows a potential for distribution without the complexity of managing vms and application deployments basically waiting for transactions.

The architecture will be reliant on kinesics streams, you can substitute various branded Kafka, but streaming is a must as you are trying to increase your footprint much wider than datacenter - Sacramento to datacenter SF to datacenter LA for instance. Think dc construct Africa to Emea, to w Europe to e Europe at a very high level.

Essentially you’re using the edge networks and internal private switching AWS maintains for themselves as well as the Amazon store, yes those are separate philosophically.

Transactions are streaming to a quantum ledger db, I’m not going to get deep into ledgering databases, but the idea is immutability at high transactional depth. Then we get the pass thru that’s ordered to ddb as the key value store to essentially validate that last chain transaction before cycling back out.

The linked doc is meant to be a teaser, not a blueprint.

Let’s say that over the past few years I’ve found myself delving more into a traditional CS pursuit as staying just a server jockey wasn’t fulfilling.
There has been more interest personally getting away from physical gear into build concepts that exceed the limits of what I used to work on.

Yeah, spending all day in VS code can suck.
No, talking to my peers that haven’t started to move on is very telling that they’re stuck in a mindset that is very easy to step over conceptually.
Learning statistics at a high level makes a lot of the hand wringing and snark about data analytics amusing, bc they’re wrong fundamentally.
Dealing with data structures and algorithms in a gear head sense means I’m bench racing code finding optimal infrastructure solutions.
I love being that guy in a meeting with a thinkpad and an m1 air in my bag and telling them “hold up, lemme run that, bc you’re off by like 90%.
 
Back
Top