Design Observability for LLM Apps

Hosted by Dr. Aki Wijesundara

Thu, Jan 22, 2026

10:00 PM UTC (30 minutes)

Virtual (Zoom)

Free to join

81 students

Invite your network

Go deeper with a course

AI Engineering Bootcamp & Certificate
Aki Wijesundara, PhD and Manu Jayawardana
View syllabus

What you'll learn

How to choose metrics that reflect real LLM behavior

Learn which signals actually matter, including quality, latency, cost, and failure modes.

How to instrument LLM apps beyond basic logs

Understand how to capture inputs, outputs, and intermediate steps without over-instrumenting.

How to detect and diagnose production issues

Learn how observability helps identify hallucinations, regressions, and silent failures.

How to use metrics to guide iteration

Learn how to connect observability data to product and engineering decisions.

Why this topic matters

Many LLM apps fail quietly in production. Without the right metrics, teams struggle to spot quality issues, rising costs, or degraded user experience. This session focuses on designing observability for LLM apps, covering which metrics matter, how to instrument systems responsibly, and how to use data to detect issues early and guide product and engineering decisions.

You'll learn from

Dr. Aki Wijesundara

AI Founder | Educator | Google AI Accelerator Alum

Aki Wijesundara is an AI researcher, educator, and founder with a PhD in Machine Learning and experience at Google's AI Accelerator, where he contributed to advancing real-world applications of machine learning and large language models. As Co-Founder of Snapdrum, he helps organizations design, build, and deploy applied AI systems that move beyond prototypes to production-scale impact.

Aki has taught over 5,000+ engineers, founders, and product leaders worldwide how to build production-ready AI systems using LangChain, RAG pipelines, and MLOps frameworks.

He serves as AI Advisor to the United Nations, contributing expertise on responsible AI implementation and policy.

His work bridges cutting-edge research and practical implementation across text, voice, and multimodal AI.

Alongside Snapdrum, Aki serves as CTO of Krybe, an AI voice automation startup, and CTO of Tilli, an edtech company redefining how learners engage with AI-powered education.

Previous students from

Google
Meta
NVIDIA
OpenAI
Amazon Web Services

Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.