Adityakrishnan is an AI Infrastructure Engineer at Simplismart, where he builds and runs the infrastructure behind large-scale GPU workloads and production LLM serving. He started out in software development and machine learning, and his work has steadily moved down the stack. He previously worked at EY.
Day to day he works with Kubernetes, Helm, ArgoCD and KEDA across AWS, Azure, GCP and on-prem clusters. He runs model serving on NVIDIA GPUs such as the H100, B200 and L40S, and builds observability with Prometheus, Grafana, Loki and NVIDIA DCGM. He is especially interested in LLM inference engines like vLLM, distributed inference and performance tuning.
As a mentor at Maestro, Adityakrishnan helps learners understand how systems behave all the way from the application down to the GPU, network and infrastructure underneath. He’s happy to help with:
MLOps and getting machine learning models into production
Kubernetes, Docker and cloud infrastructure
LLM serving, GPU scheduling and inference performance
Monitoring, debugging production incidents and root-cause analysis
Data structures, competitive programming and low-level design interviews
Starting a career in AI infrastructure, DevOps or platform engineering
His favourite question about any system is what’s really happening under the hood, and he brings that curiosity to every session.