About NeuroSpark
NeuroSpark builds and operates a high-performance AI inference platform that helps enterprises run large language models faster, cheaper, and at scale. Inference infrastructure is the foundation the entire AI application layer runs on — every AI product ultimately depends on how fast, how reliably, and how affordably models can serve their users. Our vision is to make that layer so efficient that compute is never the reason a good AI product fails.
About the Role
As an AI Inference & Training Framework Intern at NeuroSpark, you will work on systems that run, scale, and optimize modern AI models. Projects may involve model serving, distributed training, scheduling, caching, communication, framework integration, or runtime performance. This internship is a strong fit for students interested in ML systems, distributed systems, or AI infrastructure. Prior full-time industry experience is not required.
What You’ll Do
- Develop and benchmark inference or training framework components.
- Run and integrate models using PyTorch, JAX, vLLM, or related frameworks.
- Analyze latency, throughput, memory usage, and distributed performance.
- Experiment with batching, caching, scheduling, and parallel execution.
- Build multi-GPU inference or training prototypes.
- Develop testing, profiling, and developer tooling.
- Debug performance and correctness issues in real AI workloads.
- Work directly with engineers on production-relevant infrastructure problems.
Minimum Qualifications
- Currently pursuing a Bachelor’s, Master’s, or PhD in Computer Science, Computer Engineering, or a related field.
- Strong programming ability in Python, C++, or both.
- Coursework or project experience in distributed systems, operating systems, machine learning systems, networking, or parallel computing.
- Familiarity with at least one modern ML framework.
- Comfortable developing in a Linux environment.
- Strong interest in understanding how large-scale AI systems work under the hood.
Preferred Qualifications
- PyTorch, JAX, TensorFlow, vLLM, or SGLang experience.
- Experience running models across multiple GPUs.
- CUDA or GPU programming exposure.
- Familiarity with distributed training or inference.
- Experience with Docker, Kubernetes, cloud infrastructure, or networking.
- Research, personal projects, or open-source contributions related to ML systems.
