Aayush Gupta

I am a Software Engineer who builds reliable, high-performance distributed systems. My passion lies in solving complex infrastructure challenges, with a current focus on creating the platforms and tooling required to serve machine learning workloads at scale. My foundational experience comes from Flipkart (Walmart), where I specialized in extreme-scale system optimization. I scaled a geo-replicated, high-contention transactional database workload by over 100x and tuned Elasticsearch clusters for sub-50ms latency under massive load. This work taught me how to deliver reliability and performance under the strictest SLOs. At NetApp, I've applied these systems principles to the ML domain. I am a core contributor to the company's AI Data Services platform, where I've designed and shipped production-grade Kubernetes operators that automate the entire lifecycle of AI/ML services. This platform, which my work helped form the foundation of, contributed to a strategic partnership with NVIDIA. I am now channeling this experience toward the challenges of LLM serving and inference. My current projects and research at CU Boulder focus on storage layer optimizations and advanced KV-caching strategies to improve the performance and efficiency of large-scale inference workloads. Technical Specialties: Distributed Systems, ML Systems (Inference & Serving), Cloud-Native Infrastructure (Kubernetes, Docker), Go, Python, High-Performance Data Systems (MongoDB, Redis, Kafka, Elasticsearch), Storage Optimization, Linux.