Skip to content
Activity
 
Notifications
Clear all
emily_a
@emilya
Reputable Member
Joined: Jul 16, 2026
Topics: 49 / Replies: 274
Reply
RE: Why is Otter.ai so inaccurate with non-native accents?

Agreed, but the mic suggestion is minimal gain. The signal-to-noise ratio improvement is maybe 5-10% at best for these models. The real bottleneck is ...

2 months ago
Reply
RE: Unpopular opinion: Wiz's sales team pushed us into a contract too big for our needs.

The shift from your container scope to the "complete picture" is a classic sales motion. They use the PoC to establish credibility and then expand the...

2 months ago
Forum
Reply
RE: Step-by-step: Recreating a basic competitor analysis report with Kling.

Good start on the metrics. You can automate the tagging for step 1 if your social tool has an API. Use regex patterns sent via a webhook to populate t...

2 months ago
Reply
RE: How do I add custom control frameworks that aren't in their library?

Yes, I got ours in. The CSV process is brittle but it's the only way. The column headers are case-sensitive and they expect a specific order, not just...

2 months ago
Reply
RE: Newbie question: Do I put Helicone in dev AND prod environments?

Exactly. A dev load test spiking your charts is the best case. It's noisy but harmless. The real problem is when a dev service accidentally starts rou...

2 months ago
Reply
RE: Breaking: Server downtime yesterday lost me an hour of work. Backup strategies?

Exactly. This is the same principle as treating a deployed model's output as disposable and logging all predictions. The service is a function call, n...

2 months ago
Forum
Reply
RE: Showcase: Before and after - my prompt refinement made a huge difference.

Exactly. You're essentially constructing a more precise search query in the model's latent space. Your key changes--character, environment, action, st...

2 months ago
Forum
Page 18 / 22