﻿WEBVTT

00:00:00.020 --> 00:00:03.567
[music playing]

00:00:03.674 --> 00:00:05.958
Health plans
are really trying to identify

00:00:06.325 --> 00:00:10.151
examples of how AI can help them
solve their business needs.

00:00:10.453 --> 00:00:13.767
Ultimately, if you can find
the root cause and the problem,

00:00:13.834 --> 00:00:15.182
the innovation comes naturally.

00:00:15.592 --> 00:00:16.291
I'm Bill Maly.

00:00:16.358 --> 00:00:18.699
I work in our Payment Integrity
business here at Optum.

00:00:20.253 --> 00:00:22.051
The interesting thing
about technology

00:00:22.118 --> 00:00:23.241
and the value that we find

00:00:23.308 --> 00:00:25.565
is when you can bring in
both that health care expertise,

00:00:25.632 --> 00:00:27.600
together with
the advanced technology,

00:00:27.832 --> 00:00:29.973
to build very powerful solutions.

00:00:30.143 --> 00:00:32.869
Sometimes a standard rule
or an edit will do the job.

00:00:32.955 --> 00:00:34.786
But there are areas
of payment integrity

00:00:35.042 --> 00:00:36.685
where AI can add value,

00:00:36.752 --> 00:00:40.295
and that's really around areas of
driving operational efficiencies

00:00:40.362 --> 00:00:41.845
and reducing provider abrasion.

00:00:42.056 --> 00:00:44.167
Optum has been using AI
in our solutions

00:00:44.234 --> 00:00:45.168
for more than a decade.

00:00:45.311 --> 00:00:47.758
Currently, we evaluate
millions of claims per day

00:00:47.825 --> 00:00:50.382
running through
our AI algorithms and models.

00:00:50.870 --> 00:00:51.904
Over those many years,

00:00:51.971 --> 00:00:54.236
billions of claims
have been evaluated,

00:00:54.303 --> 00:00:56.360
and that has allowed us to improve

00:00:56.737 --> 00:00:58.880
the accuracy of our detection,

00:00:58.970 --> 00:01:02.680
detecting claims that, ultimately,
edits and rules can't.

00:01:02.962 --> 00:01:04.822
Another great opportunity for AI

00:01:05.025 --> 00:01:07.472
is really in workflow prioritization.

00:01:07.539 --> 00:01:10.725
So once we identify a claim
that needs human review,

00:01:11.251 --> 00:01:14.082
we prioritize that
through our AI models

00:01:14.149 --> 00:01:17.418
to route the claims to
the most experienced reviewers.

00:01:17.485 --> 00:01:20.401
So seeing what
this technology is capable of,

00:01:20.779 --> 00:01:24.417
and knowing some of the problems
that our health plans are facing,

00:01:24.831 --> 00:01:26.918
really gets me excited about
what we can bring.

00:01:27.133 --> 00:01:31.367
[music ends]