Edge Delta

Re:Invent this year was my 9th, and it was as crazy as I've ever seen it. The pace of innovation and the mentality of every business around optimizing for AI upside while accepting manageable AI risk - it is energizing. The reality is that generative AI code/building is undeniably valuable, and today it is phenomenal at getting you a service (or code within an existing service) that is about 50% production ready. In order to be ready for primetime (scalability, reliability, observability, monitoring, remediation), AI Teammates are the missing piece. They are the future of running healthy production environments. ✅ Packed booth with 1,000+ great conversations with innovators ✅ Evening happy hour with engineers from all over the world ✅ AI Teammates talk on stage from Emrah Samdan ✅ Poker tournament with "back room" convos in their truest sense Top 3 Favorite Quotes of Amazon Web Services (AWS) Re:Invent 2025: "Wait - aren't you all the Telemetry Pipelines people? So you built a real-time data streaming solution, which is now what every real-time AI use case needs. That's pretty lucky..." - from another Founder & CEO (that sounded totes jelly), sometimes it's better to be lucky than good "Those glasses look horrible on your face" - from third place finisher of the poker tournament, who was tactically trying to convince the second place finisher to swap his Meta glasses prize. The creative jabs ultimately failed :) "Using an MCP tool to try to temporarily grab streaming data from any of the observability or security platform APIs has an extremely limited potential." - from a F100 technical fellow, it's really nice to hear validation of our approach #reinvent #observability #monitoring #otel Edge Delta Splunk Datadog Dynatrace Elastic New Relic Sumo Logic Grafana Labs

💡 Surface actionable signals from alert noise with AI Teammates: the production-ready agentic observability team that cuts through on-call chaos.

🌎 Everyone’s talking about model quality, LLM performance, and fine-tuning. Recently I've used GitHub, Cursor, OpenAI Codex, Vercel v0, Replit, and a few others, but here’s the hard truth: 80% of AI outcomes are determined before you ever touch a model. (I made up this statistic but it sounded about correct). The real bottleneck? Getting the right data in the right format, with the right context. Raw logs, fragmented telemetry, missing signals, schema drift, undefined semantics… these aren't minor nuisances. They completely derail downstream systems. If you're serious about AI, your priority shouldn’t just be "AI readiness" but also telemetry pipelines that reliably accomplish normalization and real-time structuring at the source. Edge Delta does this without shortcuts. No black boxes and you get to see all your data and which formats it's being sent to best enable AI initiatives. Great #AI starts with great data plumbing.

As AI agents become more sophisticated, leaders face a new challenge: how do you ensure these agents act responsibly? In her latest blog post, Edge Delta CISO Joan D Pepin shares a practical framework for deploying AI agents safely. She breaks down where the biggest risks live — from data exposure and privilege escalation to compliance blind spots — and how to mitigate them through defense in depth. At Edge Delta, we’ve built these principles directly into our collaborative AI Teammates. Our AI agents: ✅ Operate with read-only access by default, keeping humans in control of any state changes ✅ Run on our enterprise-grade Telemetry Pipelines, which act as guardrails to prevent sensitive data exposure ✅ Provide comprehensive, end-to-end logging for full transparency and auditability Joan’s framework outlines how to start smart, scale safely, and build the trust required for responsible AI adoption. Read the full post here: https://lnkd.in/eHTA9UJb

This Wednesday @ 5pm in Seattle at Edge Delta HQ! For everyone from vibe coders to professional engineers, this will be a free and fun session to quickly vibe code a game, deploy it to AWS, and get AI Teammates to tell you how it's working. You'll have technical help to get fully setup with: AI IDE Kiro by Amazon Web Services (AWS) Elastic Kubernetes Service (EKS) AI Teammates by Edge Delta We'll have food and drinks and lots of resources walking around and getting you unstuck (because hungry means bad vibes, and stuck means bad vibes, and there will be no bad vibes). I'll see you there, but registration closes soon - https://luma.com/u6sxpqwm Yujian Tang Fatih Yildiz Matthew Miller Matt Meier Sarah Kricheff Ragghav S Serge Shevchenko 🚀 Crystal Shao Clayton Novotney #seattle #ai

👉 Join Edge Delta’s Product team on Oct 30th, 2-2:30ET, for a walkthrough of AI Teammates, the first multi-agent AI platform engineered to share context from your data across SRE, Security, and DevOps. Register today!