Summary
Thomas Kalnik is a Principal AI Engineer with 8 years of experience building and productionizing large-scale ML and AI systems across finance, healthcare, and creator-focused products. He has led multi-node distributed training (up to 671B params) and operated multi-cloud GPU fleets, delivering latency and cost optimizations for high-throughput image and multimodal inference services. Comfortable across the stack, he designs retrieval and evaluation pipelines, LLM fine-tuning (SFT/DPO), and agentic workflows while enabling engineering teams with internal AI platforms. His background in data engineering and FHIR interoperability gives him a practical edge in regulated domains, and he’s delivered novel solutions like agentic hybrid search for clinical data and a YouTube creator Thumbnail Studio. Known for marrying rigorous infrastructure tuning (DeepSpeed, NCCL, Grafana/Prometheus) with product-focused A/B evaluations, he moves models from research to repeatable production reliably.
8 years of coding experience
7 years of employment as a software developer
MBA/MS Finance, Investment Finance, MBA/MS Finance, Investment Finance at Northeastern University
Master's degree, Computer Science, Master's degree, Computer Science at University of Illinois Urbana-Champaign
Bachelor of Arts (BA), International Relations and Affairs, Bachelor of Arts (BA), International Relations and Affairs at SUNY Geneseo
Spanish