Building @tryidem. Linguist and Language Learner.
- Tweets 77
- Following 58
- Followers 14
- Likes 46
Pinned Tweet
I gave idem my Goodreads export and explained some of the books I liked and disliked and it built a personal reading curator that remembers my feedback and now makes me a six-book literary digest every month.
Gimme any recs you have!
Here’s October:
4
1
4
178
I gave idem my Goodreads export and explained some of the books I liked and disliked and it built a personal reading curator that remembers my feedback and now makes me a six-book literary digest every month.
Gimme any recs you have!
Here’s October:
4
1
4
178
I think this use case is far more useful and interesting than the typical "talk to your AI in your target language". I think that is soulless and absolutely useless. the fact is that the majority of good learning you're going to do is through passive learning not speaking.
1
8
That's why privacy-centric agents like @tryidem are so necessary. I don't think privacy will be top of mind for Meta for a while.
Deleted Muse after seeing this post on Threads about how it told some Facebook Marketplace sellers the guy’s address and they showed up at his door
Dangerous and creepy
This would have been 1000x worse if the person was a woman
28
been using @tryidem for all my workouts the last week, having automatic interactive workout plans is insane
9
This is huge for the future of evals!
Evaluations: solved.
We’ve shipped Agent Score, a single number to track how your agents are doing over time.
Every day, it evaluates outcome, reliability, cost, speed, and safety out of the box. And you can extend it to look for any kind of signal specific to your use case.
Most eval systems are reactive and have incomplete coverage. We analyze 100% of your traces and give you actionable insights you can dispatch to your coding agents automatically.
Available now @trylatitude
14
William retweeted
pro tip for Muse, Instict or Idem users
ask your agent to see if you're getting overcharged for your phone bill
@tryidem found me a plan of 7€ for 50gb in the same company while I was paying 15€ for 30gb, took me 15min to change
1
5
118
Wanted to share a great customer success we’ve had at Latitude with Tavi.
They run 75 personalised AI agent deployments, one per customer.
As that fleet grew, debugging each deployment separately stopped scaling. 🧵
1
2
1
282
Tom Robbins, Tavi’s CTO said:
“Codex can find the exact failure trace in Latitude without me adding any other context.”
It’s a good example of coding agents becoming much more useful when they have structured access to production evidence.
1
20
Tavi now has one debugging layer across its whole agent fleet.
Next, they’re looking at turning diagnosed failures into custom signals and using fleet-level dashboards for costs, sessions and tool errors. Huzzah!
Full story:
latitude.so/customers/heytav…
20