Arvind Arikatla

evals at @JudgmentLabs
Prev ads systems @linkedin, @snap, @aws

thinking about agent infra, ads, sf/ca politics, and coffee

[Arvind Arikatla @arvarik_]

Completely unserious “research” firm. More like SlopAnalysis

[Arvind Arikatla @arvarik_]

I used Judgment on my fantasy football agent and now leading the league in points. gotta moneyball your agents

[Alex Shan @alexshander03]

Today, @JudgmentLabs opens for general access.

We believe long running agent systems will generate more exhaust data than the world can currently process.

Hidden within agent traces are rich records of how to make these systems better: what users ask for, when agents stumble, and how agents interact with the external world.

Judgment is the most performant monitoring and improvement layer for long-horizon agents.

We’re excited to see what you build!

[Arvind Arikatla @arvarik_]

@HighyieldHarry Especially considering sleeper has free read apis, super easy to make an agent for fantasy football for moneyballing. Wouldn't be surprised if they locked this down though

[Arvind Arikatla @arvarik_]

@Sirupsen

[Arvind Arikatla @arvarik_]

@sriramk feel like I've heard 8 people pitch this exact idea in the last week

[Arvind Arikatla @arvarik_]

@GreenPlusAnE This framing is part of the problem, retirees don’t want to be known as the generation holding the short end of the stick. Acknowledgment that it was a poor plan and gradual reduction of benefits across generations is a better angle

[Arvind Arikatla @arvarik_]

silver lining of the RAMageddon is that it leads to creative solutions and knowing what to skip. Reminds me of the Nintendo engineers optimizing gold/silver to fit in Kanto

[turbopuffer @turbopuffer]

binary quantization lets tpuf store & score more vectors in RAM. to keep recall high, we rescore full-precision candidates from SSD

we cut vector search latency up to 4.3x on low-mem VMs by dynamically pruning rescore candidates

less rescoring→ fewer SSD reads→ faster search

[Arvind Arikatla @arvarik_]

@felpix_ Yep, need a persistent store to track progress and something to promote discipline. Google is almost there with NotebookLM but needs a personal agent to test comprehension

[Arvind Arikatla @arvarik_]

@Birdyword The BBC does not run ads within the UK. Commercial media is incredibly tough without it, and they probably resort to this to keep readership high to justify public funding

[Arvind Arikatla @arvarik_]

@saranormous Pales in comparison to the difficulty of procurement, either regulatory or supply-side

[Arvind Arikatla @arvarik_]

Love how codex/claude task estimates are always "this will take 10 developer days". Brother, we are ripping through this in 10 minutes

[Arvind Arikatla @arvarik_]

After exactly 1 prompt I am out of my 5-hour limit

[Arvind Arikatla @arvarik_]

Just in time before I immediately switch Fable 5.5

[Arvind Arikatla @arvarik_]

@_giacomo_volpe_ > defunding the schools

Well with prop 98 they’re very well funded and just take from income tax/sales tax pool

https://t.co/LIsmiWQpFS

[Arvind Arikatla @arvarik_]

So much of 2010’s cloud culture was dominated by a type of board game enthusiast engineer. Design discussions felt like a game of Catan. The latest AWS drop had people rushing to invoke their new piece

[Tom Lynch @Tom_A_Lynch]

This is what mental illness looks like

[Arvind Arikatla @arvarik_]

@alexshander03 a trip to a costco gas station is all you need to see this is true

[Norm Macdonald @normmacdonald]

At times, the joy that life attacks me with is unbearable and leads to gasping hysterical laughter. I find myself completely out of control and wonder how could life could surprise me again and again and again, so completely. How could a man be a cynic? It is a sin.

[Arvind Arikatla @arvarik_]

@eric_seufert pretty much lol. opposite ends of the spectrum in how well monetization works

[Arvind Arikatla @arvarik_]

@cognition apple just keeps winning

[Arvind Arikatla @arvarik_]

@LinkofSunshine My criteria is that if I were a GM for one season with free pick who would I choose. I think most would say Mahomes still