Software Engineer, Coding Agent Harness Engineering - New College Grad 2026
Tech Stack
Tag name is followed by "@" symbol and proficiency level value.
About proficiency levels:
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
AI @ 3
Agentic AI @ 3
Agentic Systems @ 3
Claude Code @ 3
Codex @ 3
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Details
NVIDIA is developing coding agent workload profiles to understand how agentic systems evolve and to improve performance across its hardware and software stack. The team researches reasoning models, parallel subagents, long-horizon tool use, and inference workloads, partnering across NVIDIA to advance the coding agent ecosystem.
Responsibilities
- Build and modify coding agent harnesses to research the impact of different architectural decisions.
- Architect AI-driven systems that analyze inference workload dynamics of leading agentic harnesses under realistic conditions.
- Collaborate with teams across NVIDIA's software and hardware stack to extend co-design to the harness layer.
Requirements
- Proven track record building or modifying coding agent harnesses at scale.
- Familiarity with agent evaluation frameworks such as Harbor.
- Familiarity with inference performance metrics, including TTFT, throughput, and ITL.
- Hands-on experience with frontier coding agents such as Codex and Claude Code.
- A rigorous, scientific approach to data collection, analysis, and research.
- Bachelor's degree in Computer Science or a related field, or equivalent experience.
Preferred Qualifications
- Work experience at an AI inference or agent company.
- Experience building modern agentic or AI-backed systems.
- Ability to translate data into specific product improvements with the end user in mind.
- Open-source projects showcasing agentic AI applications.
- Master's degree in Computer Science or a related field.
Compensation and Benefits
The base salary range is USD 108,000–178,250 for Level 1 and USD 124,000–195,500 for Level 2. Compensation is determined based on location, experience, and the pay of employees in similar positions. The role also includes eligibility for equity and benefits.
Applications will be accepted at least until September 19, 2026. This posting is for an existing vacancy. NVIDIA uses AI tools in its recruiting processes and is an equal opportunity employer.