My first PyOhio introduced me to accessible Python talks, welcoming community spaces, and new professional connections. I left Cleveland with new ideas to explore and more confidence in my future as a Python developer.
In This Post
- Why PyOhio
- My Favorite Talks
- Root Access: Using Python to Keep My Garden Alive
- Building AI Agents in Python
- My Model Works Locally. Why Is Production Lying to Me?
- Failed Experiments in Vibe Coding
- Looking Ahead
- Bonus Photos
Why PyOhio
As part of growing as a new data scientist and Python developer, I searched for Python conferences within driving distance. About five hours away in Cleveland, I found PyOhio.
I had attended a few hackathons and the MWC3 conference, but I still was not sure what to expect from my first dedicated programming conference. I am very happy to report that it was a great experience.
 |
| My first PyOhio and my first dedicated programming conference. |
The talks were accessible, and I could find something that interested me during nearly every time slot. Everyone was friendly, and I made several connections that I look forward to seeing again at future events and online.
One of the most valuable moments happened during an open-space discussion. I connected with Mike Smith, who shared career advice and became a new professional connection. His advice was to focus on building my story and to reach out to people the same way we were doing at the conference.
 |
| Connecting with Mike Smith and Dinky Mishra during my first PyOhio. |
It was a good reminder that conferences are not only about listening to presentations. Some of the most useful moments can come from simply joining a conversation.
My Favorite Talks
All the speakers were wonderful, but a few talks stood out because they were clearly presented, easy to follow, and especially relevant to where I am in my learning journey.
All three keynote speakers were amazing and truly inspiring. My favorite was Carol Willing’s talk about her journey through the early technology industry and a 15-year career break. Her story about returning to technology and continuing to grow was especially meaningful to me as someone entering the field through a nontraditional career path.
 |
| Carol Willing sharing her journey. |
Dr. Juliana Novic was wonderful as well, and Christopher Neugebauer gave an incredible presentation with some of the best-designed slides of the weekend.
There were 28 talks across nine different time slots, so it was easy to find something interesting throughout the conference. When none of the scheduled talks aligned with my interests, there were open spaces where attendees could propose and join informal discussions.
Below are a few of my favorite talks.
Root Access: Using Python to Keep My Garden Alive
After talking with other attendees, I think many people agreed that Bianca Ploss’s talk about using Python to care for her plants was one of the highlights of the conference.
No one wants their plants to die, so she created a solution using a weather API, which happened to be the same Open-Metro API I used for my Weather2Go driving-risk application. The code was simple and accessible, and I understood everything she shared. I also loved her ideas for expanding the project by adding information from a plant API.
 |
| Using Python and weather data to help a home garden grow. |
More than anything, the presentation made me feel like I could return next year and give my own talk about a project I created. It showed me that a conference talk does not have to begin with an enormous or highly advanced system. A useful personal project, explained clearly, can still be interesting and valuable to others.
Building AI Agents in Python: A Step-by-Step Guide with Strands Agents
AI agents had always seemed beyond my skill level. When I encountered them at Hack Dearborn last year, I was still new to coding and found the workshops difficult to approach.Morgan Willis’s PyOhio talk made the topic feel accessible. Her AWS Strands example used simple code, clear steps, and a fun Emotional Bears fan-support bot that made me interested in trying something myself.
 |
| Building an AI agent with the AWS Strands library by Morgan Willis |
Now I am excited to experiment with my own agent and see what I can build. The talk made AI agents feel less like hype and more like an approachable tool I could learn to use.
My Model Works Locally. Why Is Production Lying to Me?
Dinky Mishra gave a valuable talk about why machine learning models can fail after being moved into production.
Most of my modeling experience has been academic and notebook-based. I have trained and evaluated models, but I have never deployed one into a real production environment. Hearing about her experiences helped me understand several problems that are easy to overlook when a model only has to work inside a notebook.
She explained five areas to check when trying to help a model run successfully in production. This was also where I learned about data drift for the first time. A model can slowly lose accuracy over days or weeks because new data no longer resembles the data it was trained on.
 |
| Five reasons models fail in production by Dinky Mishra. |
Data drift is definitely a topic I want to research more. It made me realize that building an accurate model is only one part of machine learning. You also have to monitor whether it continues working after the world around it changes.
Failed Experiments in Vibe Coding
I also had the honor of meeting Al Sweigart (AL rather than AI) the author of several books widely used by the Python community.
I started reading Beyond the Basic Stuff with Python during the car trip and picked up several useful ideas for writing better code. His talk about vibe coding was fun and energetic, with demonstrations of some amazing failures and a few near successes.
The main lesson was important: we cannot always tell the difference between code that works and code that only looks like it works.
 |
| Good software needs to do more than appear correct. |
I use AI to help me write and understand code, but I do not like the idea of accepting output without understanding what it does. The talk reinforced the importance of reviewing, testing, and questioning AI-generated code rather than trusting it simply because it produces an answer.
Looking Ahead
I am excited to attend more conferences and technology events where I can connect with the community and learn from others.
PyOhio is definitely an event I would like to attend again. The community was welcoming, the talks were approachable, and the weekend gave me more confidence that I could contribute something of my own. I would love to return with a lightning talk or a full 30-minute presentation about a completed project, perhaps one connected to dogs or nature photography.
It is also the closest dedicated Python event I have found, other than local meetups that are still around two hours away. I have started researching data-focused events as well, and there appear to be more options in Michigan. The U-M Annual Data Science & AI Summit in November may be the next one I try to attend. We will see.
I came home from PyOhio with new ideas, new professional connections, and more confidence in my ability to keep growing as a Python developer.
Bonus Photos
No conference trip is complete without new stickers on the laptop and at least one unexpected sightseeing photo.
 |
| Laptop decorated with new Python and PyOhio-related stickers. |
 |
One final Cleveland discovery before heading home. |
Comments
Post a Comment