//Convrs
August 18, 2026


I was lead motion capture and technical artist on Convrs, Smarter Reality’s VR de-escalation training platform for law enforcement. I researched and benchmarked the video-to-motion AI tools available, then worked with the engineering lead to build a production pipeline around Blender, Marionette, FFmpeg, and Live Link — carrying actor performance from raw footage through solve, cleanup, and retarget into Unreal Engine 5 Metahumans running on standalone headsets. This required custom LOD treatment particularly regarding Quest 3. Custom shader and texturing solutions had to be implemented as well to reduce draw calls. The result was a repeatable process that let a small team ship new training scenarios on a monthly cadence.