Aug 11, 2025
Webinar: The New Agent Reliability Playbook
Shohil Kothari
Head of Growth
Shohil Kothari
Head of Growth
In the new agentic era, the old observability playbook breaks and traditional methods fall short as agents move from experiments to production. When prompts, retrieval, tools, and memory collide, AI systems break in subtle, hard-to-predict ways.
Atin Sanyal, Galileo co-founder and CTO, introduces a modern evaluation framework built for agent-based systems. It’s a practical, metric-driven approach to catching and correcting failure modes early by deeply instrumenting the agent loop across tool quality, error rates, latencies, and business KPIs.
He will walk through a real-world agent powering a stock-trading workflow, showing how brittle retrieval and flawed logic lead to drift, and how improved telemetry enables fast, targeted fixes.
Join our upcoming webinar to learn:
An agent observability and evaluation playbook for building reliable AI systems
Ways to trace root causes and drive continuous improvement with hard metrics
How to layer in agent observability with minimal lift
In the new agentic era, the old observability playbook breaks and traditional methods fall short as agents move from experiments to production. When prompts, retrieval, tools, and memory collide, AI systems break in subtle, hard-to-predict ways.
Atin Sanyal, Galileo co-founder and CTO, introduces a modern evaluation framework built for agent-based systems. It’s a practical, metric-driven approach to catching and correcting failure modes early by deeply instrumenting the agent loop across tool quality, error rates, latencies, and business KPIs.
He will walk through a real-world agent powering a stock-trading workflow, showing how brittle retrieval and flawed logic lead to drift, and how improved telemetry enables fast, targeted fixes.
Join our upcoming webinar to learn:
An agent observability and evaluation playbook for building reliable AI systems
Ways to trace root causes and drive continuous improvement with hard metrics
How to layer in agent observability with minimal lift
In the new agentic era, the old observability playbook breaks and traditional methods fall short as agents move from experiments to production. When prompts, retrieval, tools, and memory collide, AI systems break in subtle, hard-to-predict ways.
Atin Sanyal, Galileo co-founder and CTO, introduces a modern evaluation framework built for agent-based systems. It’s a practical, metric-driven approach to catching and correcting failure modes early by deeply instrumenting the agent loop across tool quality, error rates, latencies, and business KPIs.
He will walk through a real-world agent powering a stock-trading workflow, showing how brittle retrieval and flawed logic lead to drift, and how improved telemetry enables fast, targeted fixes.
Join our upcoming webinar to learn:
An agent observability and evaluation playbook for building reliable AI systems
Ways to trace root causes and drive continuous improvement with hard metrics
How to layer in agent observability with minimal lift
In the new agentic era, the old observability playbook breaks and traditional methods fall short as agents move from experiments to production. When prompts, retrieval, tools, and memory collide, AI systems break in subtle, hard-to-predict ways.
Atin Sanyal, Galileo co-founder and CTO, introduces a modern evaluation framework built for agent-based systems. It’s a practical, metric-driven approach to catching and correcting failure modes early by deeply instrumenting the agent loop across tool quality, error rates, latencies, and business KPIs.
He will walk through a real-world agent powering a stock-trading workflow, showing how brittle retrieval and flawed logic lead to drift, and how improved telemetry enables fast, targeted fixes.
Join our upcoming webinar to learn:
An agent observability and evaluation playbook for building reliable AI systems
Ways to trace root causes and drive continuous improvement with hard metrics
How to layer in agent observability with minimal lift
Shohil Kothari
Shohil Kothari
Shohil Kothari
Shohil Kothari