Carlos Osuna
Carlos Osuna
About
Detail
Site Reliability Engineering | Human-centered AI-fueled observability
Monterrey, Nuevo Leon, Mexico
DevOps is no longer enough to push a company forward. At least not in the current AI fueled tech frenzy. Its attempt to merge the Operations and Development teams was just a mild success as most companies instead of merging, created a separate and opposing department for all things DevOps.
As LangChain states on their LangSmith site:
“Observability is a critical requirement for applications built with large language models (LLMs). LLMs are non-deterministic, which means that the same prompt can produce different responses. This behavior makes debugging and monitoring more challenging than with traditional software.”
So DevOps must evolve into two main branches: Production Control and Observability. You must protect your Production Environments from malignant code introduced by Vibe Coding, and you must increase the observability of traditional code and execution engines (traces, logs, metrics), while at the same time opening the observability umbrella to the fledgling ecosystem of Chatbots, LLMs and GenAI.
Tools like Kong or Portkey can assist you creating AI gateways between your legacy systems and complex inference tools in the cloud but you must not leave the interactions unsupervised. N8n workflows look tempting with no-code, but also need a vigilant eye to avoid hallucinations running the experience and pumping AI API usage to the roof.
I can help you with this and also with the traditional DevOps and SRE tasks including CI/CD (including LLMs deployments), Infrastructure as Code, Grafana Dashboards, Automation and Incident Response with Major Incident Response Plans (MIRPs) that align with operational SLAs and business risk tolerances.
Contact Carlos regarding:
work
Full-time jobs