building https://xitter.cf/t.co/T9YZWu3n99
Toronto
Joined February 2008
- Tweets5.1K
- Following1.9K
- Followers3.5K
- Likes13.4K
Pinned Tweet
Today we're launching Surfaces on Traces
A new way to visualize your agent session activity, for everything from skills used, token consumption, security issues and much more.
Available for all your sessions on traces.com.
Tarun Sachdeva retweeted
Which skills lead to PRs? The traces know.
What if you could see a heatmap of how different skills are being used across your team? Are they actually helping or just taking up space?
Soon, only on traces.com:
What if you could see a heatmap of how different skills are being used across your team? Are they actually helping or just taking up space?
Soon, only on traces.com:
good sign when you keep opening your homepage to play with the animations
(from the new traces.com)
Can't stop thinking about this. Now I want a custom "IDE" for almost every type of software development ("IDE" in quotes, because I'm not even sure what to to call that tool).
I want a tool to help me make decisions on frontend component structures, on database design, on every part of the stack I can control. Something that helps me get to the best version of whatever software I'm building. And of course models will write the code, but I want to learn too!
We've been working on building trace intelligence across all your company sessions. We're building dimensions of trace data especially relevant to the offices of a CTO, CISO, and CFO. You'll be able to answer things like "how often do files in your codebase show up in agent read operations and how much does it cost" (an example visual below)
This is a fun but gnarly data engineering challenge. But IMO the real story is all the software we are creating along the way. We're building internal tools to evaluate schema choices, we use data from our agent to evaluate the types of queries to optimize for (using Jev and Luna to classify), we build visual profiling tools test query performance throughout. Making software now is so much more fun and interesting.
Anyways, I can't wait to show this stuff off. Your agent trace data contains multitudes and you need many ways to look at it. Also, if this is interesting enough that you might want to run a pilot, please DM!
We've been working on building trace intelligence across all your company sessions. We're building dimensions of trace data especially relevant to the offices of a CTO, CISO, and CFO. You'll be able to answer things like "how often do files in your codebase show up in agent read operations and how much does it cost" (an example visual below)
This is a fun but gnarly data engineering challenge. But IMO the real story is all the software we are creating along the way. We're building internal tools to evaluate schema choices, we use data from our agent to evaluate the types of queries to optimize for (using Jev and Luna to classify), we build visual profiling tools test query performance throughout. Making software now is so much more fun and interesting.
Anyways, I can't wait to show this stuff off. Your agent trace data contains multitudes and you need many ways to look at it. Also, if this is interesting enough that you might want to run a pilot, please DM!
Incredible use of Jev by @Sahaj0312 demo’d at PROGRAM last night!
Check out Not Snake: notsnake.com and how he built it: traces.com/sahaj0312
porgram (jev edition) by @tarunsachdeva at New Stadium was sickkkk. Built notsnake.com/ which is basically Snake... but not :) Check it out! If you wanna see how it all came together in like 15 mins: traces.com/sahaj0312
PROGRAM: JEV presenters are set!
➡ @sarimrmalik from @RubricLabs will show multiple practical examples of Jev
➡ @SrihariSriraman from traces.com will talk about how we use Jev in production
We'll supply API keys + pizza, just bring your laptops and ideas.
We're almost at cap now, but can get on the list here: luma.com/2rl4eoeb
If you've registered and can't make it, please update your registration. We'll take people off waitlist if anyone cancels.
Truly wild that this happened so early in the AI adoption cycle
We’re still at single digit % of households with a paid AI sub and coding agents have been around only a couple of years
Tarun Sachdeva retweeted
A few weeks ago, we launched Surfaces on Traces, a new way to visualize your agent sessions.
Today, we're launching a big update - anyone can now create a surface.
Surfaces can be just for your org, or anyone on traces.com - and are super easy to add!
Increased the capacity and only a few spots left for PROGRAM - Go grab ‘em if you’re in Toronto on October 7!
Ok feedback received this video is a touch too fast lol I’ll make a slower one and post it later today!
A few weeks ago, we launched Surfaces on Traces, a new way to visualize your agent sessions.
Today, we're launching a big update - anyone can now create a surface.
Surfaces can be just for your org, or anyone on traces.com - and are super easy to add!
A few weeks ago, we launched Surfaces on Traces, a new way to visualize your agent sessions.
Today, we're launching a big update - anyone can now create a surface.
Surfaces can be just for your org, or anyone on traces.com - and are super easy to add!
Every surface is versioned and sandboxed HTML, and your agent will have no problem creating one using your traces, or public traces.
We're adding more capabilities to make them more functional (classification is high on that list) - stay tuned for a lot more here!
If you're open to it, share a surface you've built publicly and tag @tracesdotcom.
Can't wait to see what you build!