Workshops ... Appendix: file inventory
What to improve next

Appendix: file inventory

In the public workshop repo we keep two Python projects.

The ingestion project contains these files:

flow/
├── activities.py
├── create_index.py
├── main.py
├── notebook.ipynb
├── pyproject.toml
├── utills.py
├── uv.lock
├── worker.py
└── workflow.py

The agent project contains these files:

agent/
├── agent.ipynb
├── agent.py
└── tools.py

The repo also includes:

data/
├── D2rw52SOFfM.txt
├── ...
README.md
pipeline.md
temporal-install.md
workflow.py

Here data/ holds the cached transcripts, pipeline.md holds an early plain Python version of the ingestion workflow, and temporal-install.md holds Temporal CLI install notes for multiple platforms.

Learning in public

After you build the project, share what you learned. You don't need a perfect announcement, so write it down while the details are fresh.

A LinkedIn-style post can be short:

I completed the Deep Research Agent workshop with Temporal.

I built an end-to-end AI research agent that:
- fetches transcripts from YouTube videos,
- indexes them in Elasticsearch for search,
- uses Pydantic AI to conduct multi-stage research,
- runs ingestion and agent workflows durably with Temporal.

The main things I learned:
- how to build reliable workflow activities,
- how to make transcript ingestion retryable,
- how to give an agent search and summarization tools,
- how Temporal helps long-running AI workflows survive failures.

Code: <your repo link>
Workshop materials: https://github.com/alexeygrigorev/workshops/tree/main/temporal.io

A shorter post can focus on the system:

Built a durable AI research agent over podcast transcripts.

The system:
- downloads YouTube transcripts,
- indexes them in Elasticsearch,
- searches with a Pydantic AI agent,
- summarizes long transcripts with a sub-agent,
- runs through Temporal for retries and durable execution.

Code: <your repo link>

Additional resources

These references are useful:

Workshop code

Compare your files against the final code in the workshop repo:

What to improve next

Questions & Answers

Sign up to ask questions, track your progress, and get access to other workshops · Already have an account? Sign in