ShipTalk - SRE, DevOps, Platform Engineering, Software Delivery · By Harness

Shipping Practical AI: How to Build Real-World ML for 2D Drawings (with Marina Petzel)

·45 min·4 clips
Marina Petzl explains why 70% of an ML engineer's time is spent on data cleaning, not building advanced models.
Marina Petzl, a Senior ML Engineer at Autodesk, discusses practical AI applications, including her work on automating CAD workflows with machine learning. She shares insights on data preparation, AI safety, and the future of multi-agentic systems, while encouraging hands-on learning and responsible adoption.

As heard by us

Practical AI here means better data, safety checks, and discipline around shipping.

Ship Talk treats practical AI as something teams have to deliver, not just discuss. Marina Petzl, Senior ML Engineer and AI Productivity Lead at Autodesk, keeps the focus on customer-facing ML, data curation, safety checks, and the plain fact that stronger inputs often matter…

Read the full review in PlayNext →

Why you'd press play

If you want the reality check behind shipping AI, start here.

Read the full recommendation in PlayNext →
Listen to the show on