# Day 2 Lecture: How to link your AI agent to value 

> In this lecture Zach dives into end-to-end AI applications, discusses what AI excels at, the roadblocks that hinder its value, and strategies to overcome those challenges. He shares that around 60–70% of AI projects fail, slightly better than the 80% failure rate often cited for data science projec…

- Web page: https://www.dataexpert.io/lesson/building-end-to-end-ai-application-day2-lecture-july25-p737-l1190
- Program: [AIExpert](https://www.dataexpert.io/program/ai-expert)
- Module: Week 5: End-to-End AI Applications
- Access: Requires enrollment in AIExpert
- Length: 37 min video
- Skills: NLP, LLMs
- Academy: DataExpert.io Academy

## About this lesson

In this lecture Zach dives into end-to-end AI applications, discusses what AI excels at, the roadblocks that hinder its value, and strategies to overcome those challenges. He shares that around 60–70% of AI projects fail, slightly better than the 80% failure rate often cited for data science projects. Zach explores key AI use cases—prediction, detection, and optimization—while emphasizing that not every problem requires large language models.

## Navigation

- Previous lesson: [Day 1 Lab: Using MCP to build more powerful agents](https://www.dataexpert.io/lesson/building-end-to-end-ai-application-day1-lab-july25-p737-l1188.md)
- Next lesson: [Day 2 Lab: Building an asynchronous vs synchronous agent](https://www.dataexpert.io/lesson/building-end-to-end-ai-application-day2-lab-july25-p737-l1191.md)