Unreal Engine Streaming Avatar Integration
Budget: $15 – $25 USD
I have an Unreal Engine project that presents a realtime 3-D “teacher” avatar. All core assets are ready; what I need is someone who can knit everything together so the avatar performs flawlessly on both Web and Mobile builds.
The job breaks down into three connected pieces:
• Avatar functionality
Blend existing Blueprints with targeted C++ nodes so the character lip-syncs to incoming audio, drives facial morph targets, plays contextual gestures, and calls an external AI API for dialogue responses. The Blueprint layer must stay clean so designers can tweak behaviours without touching code.
• Live streaming
The finished scene has to reach end-users through BOTH WebRTC and RTMP. You’ll wire the render output to those protocols, handle bandwidth adaptation, and expose the usual start/stop controls so my frontend team can trigger streams from JavaScript or native mobile code.
• Build, debug, deploy
Final output must ship as an in-browser build (WebGL / Pixel Streaming) and as mobile binaries (iOS & Android). I expect you to chase down platform-specific quirks—shader permutations, memory limits, input mapping—until each target runs at an acceptable frame-rate on mid-range devices.
Acceptance criteria
1. Avatar plays lip-sync, facial animation and gesture loops in response to live audio + AI text.
2. Single button starts a WebRTC stream; second button starts an RTMP stream; both verified on public servers.
3. Web build loads in Chromium-based browsers; mobile build passes App Store & Play Store release checks.
4. No critical errors or warnings in the output logs.
I work in Git so please be comfortable committing your changes to a feature branch and writing concise README notes explaining any engine-level tweaks. If you have shipped something similar, send a brief demo link or repo snippet—seeing a streamlined pipeline means we can kick off right away.
The job breaks down into three connected pieces:
• Avatar functionality
Blend existing Blueprints with targeted C++ nodes so the character lip-syncs to incoming audio, drives facial morph targets, plays contextual gestures, and calls an external AI API for dialogue responses. The Blueprint layer must stay clean so designers can tweak behaviours without touching code.
• Live streaming
The finished scene has to reach end-users through BOTH WebRTC and RTMP. You’ll wire the render output to those protocols, handle bandwidth adaptation, and expose the usual start/stop controls so my frontend team can trigger streams from JavaScript or native mobile code.
• Build, debug, deploy
Final output must ship as an in-browser build (WebGL / Pixel Streaming) and as mobile binaries (iOS & Android). I expect you to chase down platform-specific quirks—shader permutations, memory limits, input mapping—until each target runs at an acceptable frame-rate on mid-range devices.
Acceptance criteria
1. Avatar plays lip-sync, facial animation and gesture loops in response to live audio + AI text.
2. Single button starts a WebRTC stream; second button starts an RTMP stream; both verified on public servers.
3. Web build loads in Chromium-based browsers; mobile build passes App Store & Play Store release checks.
4. No critical errors or warnings in the output logs.
I work in Git so please be comfortable committing your changes to a feature branch and writing concise README notes explaining any engine-level tweaks. If you have shipped something similar, send a brief demo link or repo snippet—seeing a streamlined pipeline means we can kick off right away.