Searching...
14 results for “ai roi”
Concept2 mentions across episodes
ROI
competing insurance companies with your own data. So normally for these cases, you would have a smaller fine tune model that's fine tuned on your proprietary data, distilling your proprietary human knowledge into the models. That human underwriter th
16:30 / 17:37
from transcript16:30 – 17:37
of cases, right, we're seeing these failure cases. And the answer is that we'll need evals for every one of those implementations and examples because evals are the way that we measure the truth, that we have a stasis point of understanding what the
35:09 / 36:20
from transcript35:09 – 36:20
The number one thing is bridging the divide in the real sim gap. Like, how do we make sure that the tasks that we're building evals over and hill climbing as closely as possible reflect the distribution of the capabilities that people care about? So
26:57 / 28:04
from transcript26:57 – 28:04
You wanna make sure you're collecting data the right way. The way, for example, we do deployments is like a tandem system where you'd have a human and AI doing the same job for a period of time, where a manager can see the output of both. If the agen
31:29 / 32:41
from transcript31:29 – 32:41
Because if you think about compute as I inject a certain number of dollars and I get a certain performance back, If I use better data, then I will get more performance back per dollar invested, and now my compute is more valuable. So that's where tra
57:42 / 58:54
from transcript57:42 – 58:54
...model. And when you're building an ROI model, there are many buckets to focus on, but there are three that are very critical. The first one is, you know, what are the incremental gains that you actually bring to the table based on KPIs and metrics th
19:26 / 20:52
from transcript19:26 – 20:52
subjective, and only getting more competitive. And that's why teams like Eleven Labs, Brex, Replit, Deal, and 5,000 other organizations use MetaView, the AI company giving high performance teams a real unfair advantage in hiring. MetaView's built a s
1:04:32 / 1:05:39
from transcript1:04:32 – 1:05:39
to calibrate certain things, like how well does the model do on building an LBO model, for example. And you're gonna see more and more benchmarks cited. Now the complexity then becomes if you move from five main benchmarks, like, SuiteBench and other
42:37 / 44:07
from transcript42:37 – 44:07
Where are you not meeting the bar? Where are you not great and you are aware of it? So I think one area where I'm not great, which is kind of funny, but, one area where I'm not great is I'm really bad at understanding financials. So sometimes people
50:52 / 52:23
from transcript50:52 – 52:23
Because with AI products, you're actually bringing a lot of value to the table. And if you don't capture that from day one, then you're training your customers to expect more for less. So for instance, think about this. If you're building a, you know
28:22 / 29:31
from transcript28:22 – 29:31
They build realistic RL environments, next generation data quality systems built from real world operational traces, and coding datasets that stress models under conditions where failures matter, state changes, workflow branching, brittle tool calls,
1:05:53 / 1:06:42
from transcript1:05:53 – 1:06:42
You know, they they want they want train better. So train faster, usually doesn't matter so much from the perspective of, hey, this model is now a lot cheaper. It does matter a lot more from the perspective of you can iterate much faster. Right? Beca
56:53 / 58:10
from transcript56:53 – 58:10
insight into the quality, durability of this revenue. There's not a single good example that we can find of sustained positive margin expansion and impact of AI inside of a true corporate enterprise that is not right now a small test. There's not. So
32:14 / 33:15
from transcript32:14 – 33:15
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.