Datadog Developer

the people have spoken... a curated sticker collection is king 👑

I keep noticing a gap between how companies talk about AI agents publicly and what's actually happening inside the ones furthest along, which is why I'm hosting a panel on it at the OpenAI office in NYC on September 9th! Publicly, it's still mostly "should we adopt agents." Inside the companies actually running them at scale, the questions are different: How do you manage token costs when agents scale exponentially? what governance gaps emerge when usage goes from a handful of experiments to thousands of agents in production? What infrastructure ends up getting built in-house vs buying from a vendor? Credal is bringing together three people trying to solve these problem at the top tech companies in New York: - Head of Applied AI @ Ramp, Veeral Patel - one of the voices behind Ramp's increasingly public approach to AI infrastructure, co-authoring the “Token Casino” and the economics of AI inside the enterprise, and most recently helping bring Ramp Router from internal infrastructure into a product of its own. - Chief Technology Officer @ Flatiron Health, Allison Candido - who’s overseeing Flatiron's shift towards AI-powered, agentic workflows capable of operating across millions of patient records, and navigating what it takes to deploy AI into one of the hardest, most regulated environments of all: healthcare. - Director of Engineering & AI @ Datadog, Diamond Bishop who sits at the intersection of internal AI adoption, Applied AI, and product, led the development of the internal agent-native office at Datadog, and works on overall company enablement and upskilling of AI. We're keeping the group curated for people already doing this work inside their own companies. If that's you, request an invite in the comments!

On Aug 26, we hosted the 21st Japan Datadog User Group Meetup in Ebisu, Tokyo. We heard 7 insight themes, some focus are how AI-driven operations are already being in real business. Huge thanks to Wantedly for venue sponsor, such a inspiring space. Total 65 participants (32 onsite, 33 online) https://lnkd.in/gx78Jeyn #datadog #jddug

"we can just have the ai add logging" - the problem is that you need to redeploy to keep adding logs, and redeploying takes ALOT of time (CI, reviews, deployment, progressive rollout, …). Agents run much faster than redeploys. What if you didn’t have to redeploy? This happens try it out (MCP coming soon) https://lnkd.in/gSvrxTw9 👇

Datadog is now one connection away in Perplexity’s Agent API. With Connectors, agents can access Datadog alongside GitHub, Slack, and Google Drive - without passing a server URL or token on every request. An API Group admin connects the service once, and everyone in the group can use it. Connect Datadog → https://lnkd.in/gvsGuFGg