ML Architect

Blue Signal Search; United States; On-site; USD 180,000 - 280,000; Full time

Benefits offered by this job

Health insurance

Dental insurance

Life insurance

401(k) plan

Job summary

Blue Signal Search is seeking an ML Architect to take hands-on ownership of large-scale model architecture, distributed computing, GPU performance, and production-ready AI systems.

You will lead architectural decisions, optimize training and inference on GPU clusters, and drive the development of dependable ML platforms across data, training, evaluation, and deployment.

Qualifications

• Advanced knowledge of AI model development and production-ready ML systems.
• Experience designing scalable architectures for large models and data pipelines.
• Proven ability to optimize memory, latency, and throughput on GPU clusters.

Responsibilities

• Define and evolve production ML architecture across data, training, evaluation, inference, and deployment.
• Engineer scalable distributed training and inference on GPU infrastructure.
• Lead GPU and serving optimization for memory, quantization, mixed precision, and throughput.
• Build dependable ML platforms with strong observability and reliability standards.
• Own ambiguous ML initiatives and balance quality, cost, and safety.
• Collaborate with research and engineering to translate research into scalable production experiences.
• Develop robust data and evaluation workflows for model quality and experimentation.

Skills

ML architecture

GPU optimization

Distributed training

Production ML systems

PyTorch/JAX

Technical leadership

Model evaluation

Job description

Our client is building advanced AI systems designed to move sophisticated machine learning capabilities from experimentation into dependable, real-world production. They are seeking an ML Architect to take hands on technical ownership of large-scale model architecture, distributed computing, GPU performance, and the systems required to operate modern AI reliably at scale. This is an opportunity for a deeply technical ML leader to solve difficult zero to one challenges, influence engineering standards, and help shape how next generation intelligent products are built.

This Role Offers

• High impact architectural ownership across the complete machine learning lifecycle, from data and training through production inference.
• A highly technical environment where individual contribution, sound judgment, and rapid experimentation are valued.
• Competitive cash and equity compensation with health, dental, life insurance, and 401(k) benefits.

Focus

• Define and evolve production ML architecture spanning data preparation, model development, evaluation, inference, deployment, and operational feedback.
• Engineer scalable distributed training and inference environments across GPU infrastructure using modern parallelization and systems approaches.
• Lead GPU and serving optimization efforts focused on memory utilization, quantization, mixed precision, throughput, latency, cost, and production capacity.
• Build dependable ML platforms and services with strong standards for maintainability, observability, reliability, and production readiness.
• Own ambiguous, zero to one ML initiatives and make pragmatic architectural tradeoffs involving model quality, performance, cost, reliability, and safety.
• Partner with research and application engineering teams to translate emerging model capabilities into stable, scalable production experiences.
• Develop robust data and evaluation workflows that support model quality measurement, responsible deployment, experimentation, and continuous improvement.
• Set technical direction across complex ML systems, resolve difficult architecture and performance challenges, and establish scalable engineering practices for other technical contributors.

Skill Set

• Advanced knowledge of AI model development, including practical experience building, optimizing, and implementing complex neural network solutions in real-world environments.
• Strong background leveraging scalable machine learning platforms and parallel computing tools to support complex model development and high-volume AI workloads.
• Demonstrated GPU performance expertise across memory efficiency, quantization, mixed precision, latency reduction, throughput improvement, and production scaling.
• Strong software engineering fundamentals with a track record of building robust, maintainable, production grade machine learning systems.
• Proven ability to own architecture end to end across data, training, evaluation, inference, deployment, and ongoing production operations.
• Principal level technical influence, including establishing engineering standards, guiding technical contributors, and solving complex architectural and performance problems.
• Strong command of PyTorch, JAX, or another modern ML framework, with the ability to work effectively across evolving machine learning technology stacks.
• Experience with technologies such as high-performance LLM serving frameworks, GPU kernels or compilers, reinforcement learning workflows, multimodal models, or large scale data processing is beneficial.