Qualifications
What You Will Be DoingDevelop and enhance the core AI agents that drive Sazabi's capabilities. Design systems focused on anomaly detection, root cause analysis, and automated debugging. Engage in prompt engineering, tool utilization, and orchestration of agents. Enhance the reliability, latency, and accuracy of AI-driven workflows. Rapidly experiment with new models, frameworks, and methodologies. Transform complex real-world production challenges into structured AI workflows. What We SeekProven experience with large language models (LLMs) in production settings (agents, RAG, tool usage, etc.). A strong curiosity about the failures of AI systems and a drive to address these issues. Ability to quickly prototype and iterate based on real-world feedback. Solid engineering fundamentals—this role involves more than just prompt engineering. Comfort in fast-paced, experimental environments. Bonus: experience with observability, debugging systems, or developer tools.
About the job
Join Our Innovative Team at Sazabi
As we approach the year 2026, the tech world faces a looming "infinite software crisis." How do we effectively support, maintain, and manage the vast surge in application development?
Our solution is Sazabi: the AI-native observability platform designed specifically for dynamic engineering teams.
Sazabi empowers teams to inquire about their production systems in straightforward language, visualize operations automatically, and identify root causes up to 10 times faster. Forget about tedious instrumentation, complex dashboard setups, and alert configurations—just get the answers you need.
We are proud to be supported by innovators from industry-leading AI companies, including Vercel, Graphite, Daytona, Browserbase, LangChain, and Replit.
About Sazabi
At Sazabi, we are redefining how engineering teams operate in the software development landscape. Our innovative AI-native observability platform enables teams to streamline their processes, making it easier to visualize and address production issues without the burden of complex setups.