Senior AI/ML Inference Platform Engineer

Dialpad; Buenos Aires; Híbrido; ARS 241.615.000 - 392.625.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Competitive salary

Comprehensive benefits

Growth opportunities

Descripción de la vacante

Dialpad is seeking a Senior Software Engineer to build the shared AI/ML platform that spans training through production inference on NVIDIA GPUs in GCP. You will partner with ASR and NLP scientists to ensure reproducibility, evaluation, scalability, latency, reliability, cost, and safe releases.

The role is implementation-heavy, involving end-to-end work across training clusters, model artifacts, evaluation, and deployment pipelines, with opportunities to mentor other engineers and influence

Formación

• Seven or more years of professional software engineering experience in production systems.
• Experience building or operating ML platforms for training, inference, or lifecycle management.
• Strong backend fundamentals and production software design.

Responsabilidades

• Design, build, and improve platform capabilities for model training, evaluation, artifacts, release, and inference.
• Build and operate shared GPU training infrastructure for reliable, reproducible experiments.
• Improve scheduling, capacity management, storage, networking, observability, and accelerator utilization.
• Develop production-serving pathways for low-latency, high-throughput inference workloads.
• Integrate model-training frameworks and runtimes to meet automation and security requirements.
• Lead technical projects from design to production, mentoring engineers.

Conocimientos

Production engineering

ML systems

Python/Go

Cloud and Kubernetes

GPU workloads

Training familiarity

Applied data science

Systems & performance

Operational judgment

Technical leadership

Herramientas

PyTorch

JAX

vLLM

Triton

Descripción del empleo

Dialpad is seeking a Senior Software Engineer to build the shared AI/ML platform that spans training through production inference on NVIDIA GPUs in GCP. You will partner with ASR and NLP scientists to ensure reproducibility, evaluation, scalability, latency, reliability, cost, and safe releases.

The role is implementation-heavy, involving end-to-end work across training clusters, model artifacts, evaluation, and deployment pipelines, with opportunities to mentor other engineers and influence