ML Research Engineer, Performance Optimization

Location: Cambridge, MA, USA or San Francisco, CA, USA

Salary: $148,000–$240,000 USD annual base salary, with bonus potential and early-stage equity.

Company

hirify.global builds AI and proprietary AI Science Factory instruments to create autonomous scientific systems for discovery across medicine, materials, and energy.

What you will do

• Profile GPU workloads, develop kernels, handle long contexts, and optimize memory footprints for 1T+ parameter LLM training.
• Optimize multi-node GPU cluster performance, including collective and MoE frameworks, communication/compute overlap, RDMA, InfiniBand, EFA, and P/D disaggregation.
• Manage mid-training data pipelines across storage, streaming, tokenization, and processing.
• Optimize asynchronous reinforcement learning pipelines, including sandbox interfacing, long rollouts, router replay metadata, and weight synchronization.
• Build tested, reusable, and adaptable optimized configurations and communicate throughput/accuracy tradeoffs with researchers.

Requirements

• BS/MS in Computer Science or a related quantitative field with 4+ years of software or machine learning engineering experience, or a relevant PhD with 1+ year of work experience.
• Experience with LLM inference libraries such as vLLM and training libraries such as Megatron, including usage and throughput optimization.
• Experience running LLM inference and/or training at scale across multi-node GPU clusters.
• Familiarity with core reinforcement learning concepts.
• Proven software engineering excellence, including clean APIs, well-tested code, and thoughtful abstractions.
• Hands-on experience supporting and maintaining production libraries used by engineers and researchers.
• Strong CI/CD skills, including integration test suites, build pipelines, and automated quality gates for multi-package ecosystems.
• Passion for agentic systems and awareness of current approaches to LLM tool use and autonomy.

Nice to have

• Experience with kernel DSLs and frameworks such as CUDA, CuTe, CuTile, or Triton.
• Experience with GPU networking frameworks such as NCCL and NIXL and fabrics such as NVLink, InfiniBand, or EFA.
• Experience with distributed or compute frameworks such as Ray, Flyte, or Airflow.
• Familiarity with containerized compute backends such as Docker or Modal and sandbox orchestration.
• Experience with Ray actors, Ray Serve, or similar actor-model frameworks.

Culture & Benefits

• Full-time U.S. employees receive medical, dental, and vision coverage, employer-paid life and disability insurance, paid parental leave, and flexible time off with company-wide holidays.
• Benefits include educational assistance, commuter benefits, bike-share memberships for office-based employees, and a subsidized lunch program.
• Full-time employees outside the U.S. receive benefits tailored to their region; international salaries are set to the local market.
• Work is guided by truth, trust, curiosity, grit, and velocity in an early-stage environment focused on scientific AI.

Hiring process

• Applications from candidates are accepted; unsolicited resumes from recruitment or staffing agencies are not accepted unless the agency has been contacted directly by internal Talent Acquisition.