Evals Code Sprint: Add evaluation to your own app in one day (limited capacity)
You've been intending to add a proper evaluation system to your product for months. Swing by our office and get it done. You'll get expert instruction, workshop time, and in-person support from Weights & Biases engineers. Enter with your own production app, and leave with the ability to develop it faster than ever before. Plus, make your system more reliable, compare models in real-world use, and set yourself up for easy fine-tuning later.
This is for founders, CTOs, and working engineers who have a production application and want to add observability & evaluations. You'll need a production application and coding ability to participate. Please don't register if this isn't you.
What you'll learn:
- Implement simple tracing & observability
- Assemble a useful dataset from traces
- Create a strong AI evaluation system centered on your own needs
- Using your evaluations to guide prompting, model choice, and generally accelerate development
Who should attend:
- Founders and developers building in AI who need evaluation systems
- Teams with genAI apps in production, or nearing it
Bonus: The W&B founders will also be helping people 😎
Schedule:
9:45am: Doors open
10:20am: Presentation kicks off (high-level overview of why and how evals, best practices)
11:00am: Optional workshop content, you can also just jump into it
11:00am - 4:00pm: Adding evals to your application, with engineering support
12:30pm: Lunch
3:40pm: Optional closing presentation with learnings & impressions
4:00pm: We kick everyone out 🫡
Laptops are required. Technical experience required. Bring your own Python or TypeScript application.