Senior Engineer - AI scale infrastructure

What You Can Expect

Simulation & Modeling - Run and extend trace-driven and analytical simulations (ns-3, Astra-Sim) for AI/ML workload studies.

Performance Analysis - Profile workloads and identify compute, memory, and network bottlenecks; produce clear summaries and visualizations.

AI Workload Evaluation - Execute training, inference, and emerging agentic AI workloads at small to medium scale.

Networking Studies - Analyze collective communication performance (all-reduce, all-to-all, reduce-scatter) across topologies and fabrics.

Tooling & Automation - Build utilities for trace generation, conversion, merging, and visualization.

Prototype & Validation - Prototype distributed training and inference pipelines and validate results against simulation outcomes.

Hardware/Software Co-Exploration - Investigate emerging technologies (CXL, PCIe, accelerators, high-speed networks) and their system impact.

Scaling & Trade-off Studies - Conduct scaling experiments and performance/efficiency trade-off analyses for AI infrastructure.

Documentation & Knowledge Sharing - Document workflows and results; contribute to internal reports.

What Were Looking For

B. Tech or M. Tech in Computer Science, Electronics Engineering, or a related field

0-1 year of experience (fresh graduates encouraged to apply)

Basic understanding of computer systems (operating systems, computer architecture fundamentals)

Introductory knowledge of networking concepts (TCP/IP; interest in datacenter networking is a plus)

Familiarity with at least one simulation or modeling tool through coursework or projects (e. g. , ns-3, Astra-Sim, gem5, SST, or similar)

Exposure to AI/ML workloads via academic projects, including training or inference using PyTorch or TensorFlow

Basic awareness of modern AI workloads such as LLMs, distributed training, or agentic / multi-agent AI workflows

Programming experience in Python and/or C++ (academic or project-based is sufficient)

Curiosity and willingness to learn systems, networking, and AI infrastructure concepts

Nice to Have (Optional)

Mini-projects or internships involving simulation, networking, or performance analysis

Introductory exposure to distributed training concepts (e. g. , data parallelism, all-reduce)

Familiarity with Linux environments and simple scripting