Help Build W Talent's
AI R&D Team.
W Talent is a Malaysia-based MCN and social commerce team working across creator management, livestream shopping and short-form video. We're building an AI R&D team to develop custom avatars for hosts and creators who give us permission to use their likeness and voice in programme segments, product explainers and branded short videos.
Build the Workflow.
Shape the Content.
You'll work with our creator operations and production teams. We'll start by deploying models on our own equipment and producing prerecorded content, then explore batch production and real-time livestream interaction.
01Senior AI Applications Engineer (Digital Avatars & Local Model Deployment)
Lead model selection and on-premises deployment for digital avatars and take responsibility for the quality of the finished videos.
- Set up the model environment on company-owned GPU systems. Deploy speech generation, avatar animation, lip-sync and video generation models, and resolve software dependency and hardware compatibility issues.
- Integrate approved host likenesses, voice samples and scripts into the production workflow. Generate review-ready product explainers, prerecorded livestream segments and short videos.
- Improve likeness consistency, voice quality, lip sync and visual quality across scripts and settings. Track generation time, GPU memory use and approval rates.
- Deliver reproducible deployment instructions, model configurations and sample videos. Assess whether fine-tuning existing models would help resolve persistent quality issues.
02Backend Engineer (AI Production Systems)
Integrate deployed models into the tools our production team uses every day.
- Build workflows for entering assets and scripts, selecting avatars, generating videos in batches, previewing, reviewing, revising and exporting.
- Implement job queues, progress tracking, retries and generation history for multiple products and campaigns.
- Manage asset access, project permissions, video versions and delivery records so teams can trace the assets and settings behind each video.
- Work with the AI applications engineer to integrate model services and with the infrastructure engineer to dispatch jobs to available machines.
03AI Infrastructure Engineer (GPU Deployment & Operations)
Keep company-owned GPU systems and local model services running reliably.
- Configure machines, drivers, runtime environments and model versions at scale. Define hardware acceptance tests, upgrades and rollback procedures.
- Allocate jobs and GPU capacity across machines, reserve resources for scheduled work and manage batch backlogs.
- Monitor hardware health, GPU memory use, job failures and service availability. Manage alerts, incident recovery and production backups.
- Use actual equipment utilisation, per-machine throughput and delivery requirements to recommend how many machines to bring online and where to allocate them at each stage.
04Real-Time Audio and Video Engineer (Avatar Livestreaming)
Integrate avatar generation into a livestream workflow that production staff can control on air.
- Connect real-time speech and avatar video output to the existing production and streaming workflow.
- Improve end-to-end latency, audio and video synchronisation, frame rate and stream stability over long sessions. Test capacity under peak concurrent use.
- Work with livestream operations to support manual takeover, output pauses, reconnection and switching to backup content.
- Run controlled livestream tests and verify go-live, audience interaction and incident handling with the operations team.
05Digital Avatar Production Lead
Own authorised production assets, content standards and final video review.
- Coordinate permission and likeness and voice capture for hosts and creators. Document the permitted uses of each asset.
- Prepare product scripts, brand assets and livestream scenarios, and set delivery standards for engineering prototypes.
- Review likeness, speech, lip sync, product information and brand presentation. Consolidate feedback and follow through on revisions.
- Track production time, revision rounds and approval rates to decide where to expand the use of avatar content.
