# Automate complex workflows with OpenAI o3

## Метаданные

- **Канал:** OpenAI
- **YouTube:** https://www.youtube.com/watch?v=ydJNqND6N_Y
- **Дата:** 08.05.2025
- **Длительность:** 3:21
- **Просмотры:** 166,798

## Описание

This demo highlights how OpenAI o3 models can complete multi-step tasks and use tools—analyzing data, visualizing insights, and drafting summaries to save you hours of work.
Ideal for operations, finance, and strategy teams looking to streamline recurring workflows.

🔗 Get stared: OpenAI for Business [https://openai.com/business/]
❓ FAQs:  Help center [https://help.openai.com/en/articles/9824962-openai-o3-and-o4-mini-usage-limits-on-chatgpt-and-the-api]

## Содержание

### [0:00](https://www.youtube.com/watch?v=ydJNqND6N_Y) Segment 1 (00:00 - 03:00)

So, we've steadily improved our reasoning models, but what's most exciting about 03 is that it combines multi-step reasoning with the ability to use many tools agentically while completing its task. So, I'll go ahead and drop my prompt in here. And I'm asking 03 to run a month-end variance report based on this dummy data that I'm uploading. And if I take a quick peek at what's inside one of these many different department spreadsheets here, you can see that we have some numbers. We have budgeted and actual spends for each team as well as some other details like cost center and forecast. Now, usually when I run through a process like this manually, I would need to harmonize and analyze the data flag variances of say more than 7%. I would need to visualize that data, probably look around on the web for maybe some benchmarks that I can compare against, and then I would need to create an executive summary uh or post for my CFO. So, you can see here that there's multiple discrete tasks that I want chat GBT to go out and complete, each one using a different tool or capability. So, this is where 03 can really shine as it's going to be able to run every step of this process step by step, calling new tools as it goes. So, we'll go ahead and send this off. Now, chat GBT will start to go ahead and think, and you can actually see the chain of thoughts summarizing its actions down at the bottom of the screen. So, we have CSV files being analyzed. We have some Python code being written for the analysis of that. It says that it's going to be focusing on certain aspects of the files. And it looks like there's 25 lines. 20 are flagged indicating the most categories uh exceeding 7%. All right, next it's searching the web finding some details from citable sources. Looks like we're getting some stuff from say KPMG and we can see some analysis of the documents that are being put in and what's being turned out. So, in just about a minute, we've got some information about the flag 7% variances. A bit further down from that, we've got some data visualizations. We have an analysis of the variance with some interactive visuals. And we can click into those and actually see some different numbers here. And then there's some key takeaways if we want to dig in a bit further on the information that we have. again, citable sources across the line, a quick write up on maybe how to read those visuals, and then some checkpoints just to make sure that we're getting what we actually expect out. Then down at the bottom, looks like we have our executive summary and a Slack post ready to go to our CFO. So 03 has gone away and automated three to four steps of what's normally a very long workflow using whichever tools it needs to get the job done and generate a report that normally would have taken hours.

---
*Источник: https://ekstraktznaniy.ru/video/11308*