Research Engineer (AI)

Location: San Jose office (HQ) or New York City office, United States

Salary: $200,000–$250,000 per year

Company

hirify.global builds an AI-powered enterprise transformation platform that connects business processes, data, and code and changes complex systems in weeks rather than years.

What you will do

• Build and scale post-training systems for supervised fine-tuning, preference optimization, and reinforcement learning for long-horizon tool use and enterprise-system transformation.
• Develop memory, context, ontology, and knowledge-graph infrastructure for agents operating across complex enterprise environments.
• Design synthetic-data, data-curation, and curriculum pipelines for enterprise landscapes, transformation traces, and tool-call trajectories.
• Build sandboxed RL environments and execution-and-verification harnesses with automatically verifiable rewards.
• Own offline evaluation infrastructure for long-horizon agent behavior, including trajectory scoring, task suites, reproducibility, and experiment tracking.
• Optimize distributed training and inference throughput, then move validated results into production serving with quantization, configuration, and rollback paths.

Requirements

• Significant experience training, fine-tuning, or post-training language models with measurable results owned directly.
• RL experience in areas such as RLHF, RLAIF, RLVR, GRPO, or agentic reinforcement learning.
• Strong software engineering fundamentals and fluency in Python and PyTorch or JAX.
• Ability to design, run, debug, and interpret rigorous experiments while distinguishing real effects from noise and bugs.
• Experience with GPU infrastructure and distributed training at scale.
• Clear written communication and an interest in taking research results into production.

Nice to have

• Experience building RL environments, execution sandboxes, or verifiable-reward task suites.
• Experience with long-context modeling, agent memory systems, knowledge graphs, ontologies, or semantic layers over enterprise data.
• Experience with code models, repository-scale context, program synthesis, automated repair, or transpilation.
• Open-source contributions to ML systems such as vLLM, SGLang, PyTorch, Triton, DeepSpeed, Ray, Megatron, or TRL.
• Publications, technical reports, open-source releases, or an advanced quantitative degree.

Culture & Benefits

• Product-focused environment building a governed, vendor-agnostic enterprise AI platform.
• Research engineering owns the machinery and its operation at scale, working closely with Research Science.
• Work includes open-weight models trained on rented clusters under compute constraints.
• Emphasis on reproducibility, experiment tracking, clear ownership, and production impact.