Nathen Harvey

I wish more companies were like Salesforce in offering a sandbox env for development and testing. That said, they have to for it because nothing is straightforward when integrating vendor offerings into their application.

2

Seemingly small features can have a huge impact. With the new AI powered help chat, IT admins can get detailed insights on how to use PDQ Connect to accomplish their goals - all without needing to leave the product. Proud of what the team has delivered here!

25

In early January, we quietly added native support for Tracing with OpenTelemetry to Sematext. 🚀 Let me skip the usual blah blah about OTel and no vendor lock-in and such and jump straight to what you get. Links at the bottom. 💰50–80% lower cost than Datadog, New Relic, Dynatrace or Grafana Cloud This is self-explanatory. :) ... and may be enough for some folks to give it a go. ✅ Really dead-simple setup If you’re already using OpenTelemetry, you’re already set! Sematext Agent acts as the Collector. It exposes HTTP and gRPC endpoints. Just point your instrumentation SDK at the agent. ✅ Immediate out of the box insights For each service see RED metrics (rate, errors, duration), see slow traces, alerts, overall latency distribution, top slow endpoints, health of each service… ✅ Intuitive, SRE-friendly UI Fast, clean and modern UI. Quickly drill from service → endpoint → trace → span. ✅ Auto-instrumentation out of the box For languages/runtimes and frameworks that have auto-instrumentation just use that. For others, instrument manually. All instructions are copy/paste in the UI. Some links: - docs: https://lnkd.in/e5mh_nzW - some screenshots: https://lnkd.in/eZZs5757  - agent & OTel: https://lnkd.in/euE2nDdP - product info: https://lnkd.in/erhEK6SJ #OpenTelemetry #Tracing #Observability #DevOps #SRE #SoftwareEngineering #Microservices #Monitoring #Sematext

43

Day 2 — Modernization Is Not Migration. It’s Momentum. If Post 1 was about why data platforms feel harder than they should, this post is about moving forward without breaking things. In my conversations with CRIS, one theme was very clear: “We can’t afford big-bang cutovers anymore.” Everyone has lived this story at least once: • long “freeze” windows • weekend cutovers that spill into Monday • parallel systems nobody fully trusts • teams praying the rollback plan stays on the whiteboard That’s not modernization. That’s survival. When we dug deeper with CRIS, we explored a different model: 🔹 Modernization as a continuous flow, not a one-time event. 🔹 Near-zero downtime as a design goal, not a miracle. On the EDB PG AI Platform, that looked like: • Using a Migration Copilot to guide the journey step by step • Streaming data from existing systems into Postgres with near-zero downtime • Keeping applications online while schemas evolve • Treating upgrades and patching as routine, not crisis projects • Pulling in data from external sources without constant manual plumbing The result isn’t just a “new database”. It’s a different rhythm: • Change becomes smaller and safer • Risk becomes managed, not avoided • Teams stop arguing about when to modernize and start asking what’s next Modernization is not about leaving something behind. It’s about unlocking what comes after. With CRIS, the most energizing part wasn’t the tools themselves — it was the realization that they don’t have to choose between stability and progress. In the next post, I’ll talk about where this leads: how operational and transactional data can start powering ML models, AI agents, and support experiences built on real, trusted data. For now: Modernization isn’t a cutover date on a slide. It’s the moment your platform starts supporting continuous change instead of resisting it. #Modernization #DataPlatforms #PostgreSQL #EDB #PGAI #EnterpriseIT #TechLeadership #CRIS

13