Staff AI Systems Engineer, Computer-Use Agents

Hark; San Jose (CA); On-site; USD 180,000 - 400,000; Full time

Job summary

Hark in San Jose is seeking a Member of Technical Staff for Computer-Use Agents to join a high-performance team that blends software engineering with research. You will help build and optimize agents that use computers to complete long-horizon tasks, spanning data, environments, evaluation, and training infrastructure.

You will own environments, evaluation, post-training workflows, and data quality, collaborating across harness, product, and infrastructure teams to ship reliable, scalable AI

Qualifications

• Strong software engineering and prototyping skills.
• Ability to turn open-ended problems into crisp hypotheses.
• Hands-on experience training, fine-tuning, or evaluating large language or multimodal models.
• Deep expertise in data/environments, agent evals, RL post-training, or training/inference infra.
• Comfort with ambiguity and bias for action.
• Clear technical communication across research, infra, and product.

Responsibilities

• Environments & Data: Build and scale interactive environments and task generators for computer use.
• Evaluation: Design benchmarks and eval harnesses to measure long-horizon task success.
• Post-Training: Develop and scale SFT, RL post-training, reward modelling, and verifiers.
• Research & Novel Methods: Identify new algorithms and prototype with strong baselines.
• Data Quality: Build filtering, curation, and curriculum methods for high-signal data.
• Agent Infrastructure: Scale training and inference infra for large-scale agent rollouts.
• Production Impact: Collaborate with harness, product, and infra to ship improvements.

Skills

Engineering excellence

Research ability

Model experience

Domain depth

Ownership

Communication

Job description

Hark in San Jose is seeking a Member of Technical Staff for Computer-Use Agents to join a high-performance team that blends software engineering with research. You will help build and optimize agents that use computers to complete long-horizon tasks, spanning data, environments, evaluation, and training infrastructure.

You will own environments, evaluation, post-training workflows, and data quality, collaborating across harness, product, and infrastructure teams to ship reliable, scalable AI