WEBVTT

00:00:05.580 --> 00:00:07.914
We're presenting qubit refill. We're picking the best

00:00:07.914 --> 00:00:10.377
crypto using quantum computers. So instead of optimizing

00:00:10.377 --> 00:00:12.928
what assets you can buy, we're optimizing how you purchase

00:00:12.928 --> 00:00:15.490
those assets.

00:00:17.590 --> 00:00:21.030
This works based off of this graph based problem where you

00:00:21.030 --> 00:00:24.528
can find a ground state energy and we pick the asset that's

00:00:24.528 --> 00:00:27.794
the least well performing to optimize your selection so

00:00:27.794 --> 00:00:32.010
that you are minimizing your slippage.

00:00:32.220 --> 00:00:35.662
And the way we do that is pretty straightforward. We build

00:00:35.662 --> 00:00:38.987
the problem, we race the solvers, then we reallocate and

00:00:38.987 --> 00:00:42.371
then we emit event. So we're just running an optimization

00:00:42.371 --> 00:00:45.638
pass each time. Once we do that, we figure out what the

00:00:45.638 --> 00:00:48.905
asset is, the optimal one and then we pick it to do the

00:00:48.905 --> 00:00:52.289
purchase. And as I said, that optimizes our slippage. And

00:00:52.289 --> 00:00:55.731
over time, that's going to save the user lots of money and

00:00:55.731 --> 00:00:59.890
then we're going to jump into a demo.

00:01:06.930 --> 00:01:10.038
So in this demo, we're using the MCP server and we're going

00:01:10.038 --> 00:01:13.096
to make a purchase of those mobile game cards that you put

00:01:13.096 --> 00:01:16.153
into the Telegram and all of that's going to be automated.

00:01:16.153 --> 00:01:19.210
And as part of that process, we're going to be picking out

00:01:19.210 --> 00:01:22.112
the optimal crypto, which as we know right now, because

00:01:22.112 --> 00:01:25.065
it's been flat for the last week, is Bitcoin. If it were

00:01:25.065 --> 00:01:27.967
going up, then we would see, you know, some other asset

00:01:27.967 --> 00:01:31.076
gets selected instead. But since it's not moving as much as

00:01:31.076 --> 00:01:33.926
the other assets and it's been badly performing, we're

00:01:33.926 --> 00:01:36.960
going to get rid of that one first because it's the best.

00:01:36.960 --> 00:01:41.751
To choose if you're a finance guy and note we we left all

00:01:41.751 --> 00:01:43.230
of the.

00:01:43.230 --> 00:01:46.691
Prompts. We didn't turn on auto mode and we did that

00:01:46.691 --> 00:01:50.544
intentionally just to show you sort of where the where the

00:01:50.544 --> 00:01:54.331
thing gets stuck. But in practice, you would just turn on

00:01:54.331 --> 00:01:58.249
auto mode and it would make these selections automatically,

00:01:58.249 --> 00:02:01.841
but we're selecting what the LLM would select as we go

00:02:01.841 --> 00:02:03.430
through it.

00:02:03.430 --> 00:02:07.340
And you can kind of see the logic as it goes through. It's

00:02:07.340 --> 00:02:11.316
now selected the Bitcoin and it's giving you numbers on why

00:02:11.316 --> 00:02:16.630
it's selected that based on our optimization pass.

00:02:21.770 --> 00:02:27.740
And you can see it's going to call the bit refill at MCP.

00:02:27.740 --> 00:02:32.890
And it will make our purchase in just a moment.

00:02:44.770 --> 00:02:48.036
And now that we've bought the product, that's sort of the

00:02:48.036 --> 00:02:51.359
the whole game. So what we've done here just one more time

00:02:51.359 --> 00:02:54.512
is we've used quantum unconstrained binary optimization

00:02:54.512 --> 00:02:57.722
protocol which you can read more about on the website we

00:02:57.722 --> 00:03:00.933
set up and we are using that to select the best asset in

00:03:00.933 --> 00:03:04.312
which to make the purchase given current market conditions.

00:03:04.312 --> 00:03:05.850
And then once we do that.

00:03:05.850 --> 00:03:08.692
We're minimizing slippage overtime for the user so that

00:03:08.692 --> 00:03:11.433
they are they have more money to spend and buy things

00:03:11.433 --> 00:03:13.890
overtime.