Units are Types: Let's Treat Then That Way
(20 Minutes)
By: Emmanuel I. Obi
Experience Level: Intermediate
Slides Link

At SREcon 2026, I showed how untyped infrastructure configs fail silently and how dimensional analysis catches those failures at definition time. But catching errors in human-written configs is only half the story. AI agents are now generating configs, writing IaC, computing dosages, and scaling parameters autonomously. They make the same unit mistakes humans do, faster and more confidently.

This talk extends the SREcon argument into agentic territory. I’ll show how ucon’s MCP server turns dimensional analysis into a verification primitive that any AI agent can call not as a unit converter, but as an algebraic safety net. I'll walk through real agent workflows where structured error responses (dimensional vectors, got/expected pairs, likely-fix hints) enable a model to self-correct in a single retry, and demonstrate how kind-of-quantity enforcement catches errors that no existing unit library can even represent. The core claim: small models paired with algebraic verification outperform large models without it and the infrastructure to prove that should be a tool call, not a training objective.

You Can Start a Tech Club in Your Community
(45 Minutes)
By: Brianne Caplan
Experience Level: Novice

You don’t need a nonprofit, funding, or a perfect curriculum to start a tech club—just a starting point.

Today, access to computer science education is still deeply uneven. Only about 60% of U.S. high schools offer foundational computer science, and participation gaps persist for girls, Black, and Latino students. Further, most students are unlikely to pursue tech pathways unless they are exposed to computer science early in school and in ways that feel relevant to their lives.

In this talk, I’ll show how developers, students, and professionals are starting tech clubs in their communities using the skills they already have, and how Code Your Dreams supports people in actually doing it. We’ll walk through how to go from idea → first session, how to teach beginners (even if you’ve never taught before), and how to build something that lasts.

I’ll share real examples of clubs launched in schools and community spaces, what worked, what didn’t, and how small efforts can turn into real community impact.

If you’ve ever thought, “I’d love to give back, but I don’t know where to start,” this is for you.

What You’ll Learn:

• Why computer science education is still important in the age of AI
• How to launch a club or get involved in your community
• Ways to teach coding / tech and keep participants engaged
• How Code Your Dreams can support you in getting started

The Job Market, a Tale of Pain
(20 Minutes)
By: Hugo Seguin
Experience Level: Novice

In this talk, I share what it was really like spending seven months trying to land a job in today’s market, applying to both data analyst and data engineer roles. I’ll walk through how I approached it and how I changed my techniques along the way.

I’ll talk honestly about what actually helped me get interviews, what I found didn’t work, and the mistakes I made early on and currently. That includes how I handled resumes, technical interviews, networking, and figuring out what companies are really looking for when they hire for data roles.

I’ll also share the resources and tools I wish I knew about from the beginning things that would have saved me time, energy, and a lot of trial and error. The goal of this talk is to give practical, experience-based advice so others can approach their job search with a clearer plan and avoid some of the frustration I ran into.

Marimo: The next generation notebook
(30 Minutes)
By: Henry Cuzco
Experience Level: Novice

Many people in the data science community are familiar with jupyter notebooks and use them regularly as part of their day-to-day tooling. While jupyter notebooks have a lot of extensions and are a great way to combine documentation with code, they don't come without drawbacks. One of the biggest drawbacks is the execution order of the cells. The other drawback is looking at the code changes between commits when trying to upload notebooks to GitHub. To solve these shortcomings, marimo notebooks were created. This talk will demonstrate marimo notebooks' functionality, the modes in which you can run them, and other marimo features.

SQL Testing: Python Tools, Patterns, and Automation
(45 Minutes)
By: Paul Zuradzki
Experience Level: Intermediate

This talk is a practical introduction to SQL testing using Python. We’ll cover why testing SQL can be tricky, compare strategies for preparing a SQL test environment and preparing data, how to check for schema breakages, data validation checks, automated formatting, and automated testing in local or remote environments. Attendees will leave with patterns and code examples to make SQL pipelines more reliable and maintainable.

Bots, Androids and Academia: A Tail of Fraud Detection.
(30 Minutes)
By: Hugo Seguin
Experience Level: Novice

In this presentation, I will summarize my experience dealing with malicious actors who attempted to infiltrate my research study. I will discuss what happened, outline the stages we went through, the fixes we implemented, and our decision-making process. I will also conclude with a broader discussion of the state of behavioral science in addressing such cyber threats. More generally, I will discuss how technology companies and tech-savvy individuals can help the social sciences/Academia protect themselves from deceptive actors and raise awareness among academics about potential pitfalls.

Different Package, Same Snek?
(15 Minutes)
By: Weezel
Experience Level: Novice

You can build Python from source, use a Docker image, install it with a package manager, or use the version that comes with your operating system.  They're all the same Python, right?

Yes and no.  This talk will 1) define Python, and 2) look at what changes, depending on how you install it.   

 

Slow, Inaccurate and Brittle: How I Built The World's Worst Data Pipeline
(15 Minutes)
By: MK Barton
Experience Level: Intermediate

"There is nothing so permanent as a temporary solution". Come learn how a combination of misunderstanding, bad assumptions, tech debt and general chaos created the world's worst data pipeline.

It's slow. It's inaccurate and sadly it lived in production for 3 years. Learn from my mistakes!

Hack Finance with Wayy Research!
(20 Minutes)
By: Rick Galbo
Experience Level: Intermediate

Have you ever wanted to do some financial research, but struggled. Well maybe not, but if you're here with us tonight, maybe check out some packages that Wayy Research has been creating. Even if you're not a seasoned fin-head, you might be wondering how to get started in quantitative finance. And if you were, boy, are you in luck!

 

We will go over 3 packages:

Wayy-Research/wrdata - a financial data gathering package

Wayy-Research/fracTime - a time series forecasting library

Wayy-Research/wrtrade - a quantitative trading libarary

 

All of these will allow you to get started with quantitative trading right in the comfort of your own home. This will be a lightning round covering the basics and we will get you from zero - to - hero faster than you can say Benoit Mandelbrot. So bring your thinking caps and lets get ready to do some digging!