Omniverse & Agentic AI Engineer

Location: Madhapur, Hyderabad (In-Person) Schedule: Monday – Friday (5 Days In-Office)

Job Summary:
We’re seeking an engineer to help build interactive digital twins in NVIDIA Omniverse and wire up agentic AI workflows that plan, simulate, and automate tasks. You’ll script Omniverse Kit extensions in Python, work with OpenUSD scene graphs, prototype scenarios in Isaac Sim, and connect LLM-based agents to real-time services and APIs.

Responsibilities:

1. Develop and maintain Omniverse Kit extensions in Python for real-time simulation, visualization, and data I/O.
2. Author/manage OpenUSD (USD) stages, layers, and variants; build import/export pipelines from CAD/DCC tools into USD.
3. Prototype robotics/operations scenarios in Isaac Sim.
4. Implement lightweight agent workflows.
5. Add safety/validation to agents with NVIDIA NeMo Guardrails.
6. Integrate optimization calls when agents need fast decisions.
7. Expose twin/agent capabilities via small services and APIs, and document them for internal users.
8. Write clear demos, benchmarks, and how-tos; participate in code reviews and knowledge-sharing.

Qualifications:

1. Bachelor’s degree in CS/CE/EE or related field.
2. 1–3+ years of software experience with strong Python fundamentals.
3. Hands-on with: Omniverse Kit scripting, OpenUSD scene management, or Isaac Sim.
4. Working knowledge of LLMs and agent frameworks, and function/tool calling patterns.
5. Familiarity with adding guardrails/safety to LLM apps.
6. Comfort with Git, containers, and cloud.
7. Clear written/verbal communication; ability to turn a rough problem statement into a small prototype.

Skills:

NVDIA tech stack): Isaac Sim, Omniverse Replicator, NIM (NVIDIA Inference Microservices), Triton Inference Server, cuOpt

Basic Skills needed: Omniverse, NVIDIA NeMo Guardrails, OpenUSD