rohil gupta | Lead ML Infrastructure Engineer

Rohil GUPTA Lead ML Infrastructure Engineer Lyon, Auvergne-Rhône-Alpes, France (FR) 500 connections • 1,811 followers About As a Lead ML Infrastructure Engineer, I specialize in architecting and optimizing scalable machine learning systems, with a deep focus on MLOps, NLP, and Cloud Infrastructure. My expertise lies in bridging the gap between research and production, recently focusing on high-performance LLM and ASR deployment. In my recent work, I have driven significant efficiency gains, including automating GPU scaling (H200/H100) to cut nighttime costs by ~40% and migrating deployments to Helm charts for standardized delivery. I led the performance engineering of a batched ASR system that delivered an 18x latency reduction and 128x throughput gain, alongside optimizing a RAG engine for 5x faster performance. I am heavily invested in observability and security, having developed production-grade Python libraries for telemetry (integrating VictoriaMetrics/Grafana) and designed secure, stateless content moderation infrastructure (LLM-Guard). My background extends to healthcare, where I previously deployed complex NLP models for clinical trials. I thrive on solving real-world challenges—whether it's benchmarking H100 GPUs, optimizing JSON generation for OSS LLMs, or building robust CI/CD pipelines—to create systems that are as efficient as they are innovative. MLInfrastructure MLOps LLM RAG GenerativeAI Kubernetes GPUOptimization H100 H200 ASR CloudComputing AWS Scaleway Nebius Helm DevOps Python VictoriaMetrics Grafana Security FastAPI Automation DataEngineering Total Experience: 7 years and 7 months Experience Diabolocom(https://www.linkedin.com/company/diabolocom) Lead ML Infrastructure Engineer (Current) Apr 2026 - Present (3 months) in Lyon Department: Engineering and Technical • Level: Head Senior Machine Learning Infrastructure Engineer Jan 2026 - Apr 2026 (3 months) in Paris, Île-de-France, France