3 days ago
Dublin, Ireland +2 moreSenior
Responsibilities
- Evaluate, adapt, fine-tune, and deploy real-time speech-to-speech, full-duplex video, expressive avatar, and synchronized audiovisual models.
- Improve turn-taking, interruption handling, multilingual interaction, and responses to verbal and visual cues.
- Track research and open-source releases, benchmark approaches, and build data and evaluation pipelines for realistic conversations.
- Train and deploy models on GPU hardware across global infrastructure while optimizing streaming inference, GPU utilization, latency, serving cost, synchronization, and reliability.
- Own features end to end across model experiments, inference services, backend services, APIs, data models, and customer portal interfaces.
- Use coding agents and development harnesses to implement, test, debug, review, and maintain changes across codebases.
Requirements
- 6+ years of experience in machine learning or related engineering.
- Production experience and deep expertise in speech generation, speech-to-speech systems, video generation, or multimodal models.
- Strong Python and PyTorch skills with hands-on experience adapting pretrained models.
- Experience with streaming or latency-sensitive systems, rigorous evaluation, and practical tradeoffs among model quality, responsiveness, reliability, and cost.
- Ability to work across ML, backend, and frontend codebases and own features through a working customer experience.
- Willingness to use evolving coding agents and development harnesses while maintaining strong engineering quality and accountability.
Benefits
- Work at the intersection of AI, real-time communications, and infrastructure with access to state-of-the-art hardware and a global deployment footprint.
- Remote role, indicated by the #LI-REMOTE designation.
- Own both AI models and the products they power for customers worldwide.
Tech Stack
Categories
About Telnyx
Telnyx builds a communications platform and AI infrastructure for developers and enterprises, offering SIP trunking, voice and messaging APIs, WhatsApp Business, eSIM/IoT connectivity, and low-latency inference on a private global IP network. The privately held company, founded in 2009 and headquartered in Austin, operates worldwide and holds a Saudi VVSP license with regional anchorsites for data locality. Revenue comes from usage-based APIs and enterprise connectivity services.