Arthur Pastel

@Art049

Building @CodSpeedHQ, we make software faster for a living.

New York, NY

apas.tel

Replies Reposts Media Articles

Arthur Pastel
@Art049
Sep 5
First try of GPT Astra and now that models are smarter and finally respect skills trigger, I start to discover some really bad ideas were planted there for a long time. @AikidoSecurity I just installed the Claude plugin to use your service with my agents and now on each write
1
Arthur Pastel
@Art049
Sep 4
I don't post that much about products but honestly if you're not using @pydantic Logfire, you're missing out. I used many observability platforms, cloudwatch, datadog, baselime, grafana and it always left me with a bit of a bitter taste, constantly annoyed by the shitty
2
Arthur Pastel
@Art049
Aug 3
great to see optimized building blocks can still be added even in 2026
Daniel Lemire
@lemire
Aug 2
Article
How fast is C++26’s std::hive?
C++26 adds a new container to the standard library: std::hive. It is meant to occupy the ground between std::vector and std::list. Like a vector, it keeps its elements in contiguous blocks of memory,...
Arthur Pastel
@Art049
Jul 28
SOC 2 Type II ✅ CodSpeed runs inside your CI, so we're in scope for your security review. Controls tested over an observation window by an independent auditor, not a point-in-time check. Full report available on request. Learn more: codspeed.io/changelog/2026…
1
Arthur Pastel
@Art049
Jul 22
Frontier models have no runtime performance signal. Not in training data, not as feedback. They can’t measure the code they generate, so they guess. Exciting to talk with @heavybit about why AI-generated code needs performance guardrails, and what we’re building at @codspeedhq.
Heavybit
@heavybit
Jul 22
Does more AI-generated code mean more performance issues and software regressions? We spoke with @Art049, the founder of @codspeedhq, a performance intelligence platform built specifically to help earlier in the SDLC. heavybit.com/library/articl…