Overview
Welcome to our how-to guides for optimizing and fixing AI applications. These guides offer practical solutions to common problems across three key areas:
Each guide shows you how to spot problems, understand what metrics reveal, see examples of poor setups, and follow clear steps to fix issues.
Troubleshooting and Optimizing Your AI Applications
We’ve organized these guides by application type to help you solve specific problems in your AI systems. Whether you’re dealing with hallucinating models, inefficient agents, or poor retrievals, you’ll find practical advice based on metrics and proven fixes.
Choose the section that matches your current challenge, or explore all areas to build a better understanding of AI optimization.
Logging basics
Learn the basics of logging with Galileo.
Log Using the OpenAI Wrapper
Learn how to integrate and use OpenAI’s API with Galileo’s wrapper client.
Python
Log Using the @log Decorator
Learn how to use the Galileo @log decorator to log functions to traces
Python
Create Traces and Spans
Learn how to create log traces and spans manually in your AI apps
Python
Agentic AI
When AI agents fail to complete tasks correctly, workflows break down and users lose confidence. Our guides help you identify why agents struggle and how to improve their ability to reason and complete complex tasks.
Conversational AI
When AI chat systems give wrong information or unclear answers, users lose trust. Our guides help you fix these conversation problems using key metrics, so you can build AI that communicates clearly and accurately.
Instruction Adherence
Help your models better follow user instructions.
Python
Fixing Hallucinations and Factual Errors
Reduce made-up or incorrect information.
Reducing Hesitation and Uncertainty
Create more confident and clear responses.
Python, TypeScript
Luna 2
Learn how to leverage the Luna 2 model for scalable, real-time, customizable evaluations for enterprises.
Evaluate metrics with the Luna 2 model
Learn how to evaluate metrics cheaper and faster using the Luna 2 model.
Python
Use Luna 2 in your experiments
Learn how to use Luna 2 metrics when running experiments in code.
Python
Metrics
Learn how to create and use metrics in your evaluations.
Retrieval-Augmented Generation (RAG)
When your RAG system can’t find the right information or use it properly, answers become less accurate. Our guides help you fix these knowledge retrieval problems so your AI can provide more accurate, relevant responses.
Basic RAG Example
Learn how to implement a basic RAG system using Galileo and OpenAI.
Python
Preventing Out-of-Context Information
Keep irrelevant information out of responses.
Python
Ensuring Complete Use of Retrieved Data
Make sure your system uses all the information it finds.
Python
Maximizing Chunk Utilization
Improve how your model uses retrieved text.
Python
Fixing Irrelevant Retrievals
Make sure your system finds the most relevant information.
Python