Open to opportunities

Jayakumar Pujar

@jayakumarpujar

Message

Senior Machine Learning Engineer specializing in production-ready multimodal AI and MLOps.

United States

Message

What I'm looking for

Seeking a role that combines machine learning with infrastructure development, preferably in a research-focused environment with opportunities for technical growth and innovation.

I am a Senior Machine Learning Engineer with 5+ years of experience translating research into scalable production systems, specializing in agentic RAG architectures, geospatial computer vision, and MLOps.

I have led cross-functional teams to deploy systems handling millions of daily transactions and architected multimodal solutions (CLIP/ViT) for high-stakes domains, achieving strong accuracy and real-time inference performance.

My work spans end-to-end ML engineering: data-centric pipelines, distributed training on multi-GPU clusters, production CI/CD with Kubeflow, and cloud-native deployments using AWS and GCP.

I deliver actionable systems—vectorized GIS outputs for disaster response, decision-support tools for public health, and robust Computer Vision pipelines for Digital KYC—while reducing operational toil and labeling costs through automation and advanced model optimization techniques.

Experience

Work history, roles, and key accomplishments

Current

Geospatial ML Engineer Intern

Current

USD

May 2025 - Present (7 months)

Designed a data-centric MLOps pipeline to map the 2024 Southeast South Dakota flood using Mask R-CNN and 3m PlanetScope imagery, achieving 96.20% accuracy and 90.56% F1; delivered production-ready GIS vectorization scripts for rapid disaster response.

MLOps Python

Graduate Research Assistant

South Dakota Board of Regents

May 2024 - Dec 2024 (7 months)

Architected a modular Public Health Risk Assessment Tool for the State Department of Health using Python and Streamlit, automating data ingestion and dynamic risk scoring to visualize infrastructure gaps for state planners.

Python Streamlit Pandas Risk Scoring Health Data

Research Assistant

University of South Dakota

Jan 2024 - May 2024 (4 months)

Developed a multimodal transformer for automated radiology report generation using CLIP (ViT-B/32) and BERT, implemented distributed Slurm multi-GPU training for 80,000+ high-resolution chest X-ray pairs and integrated Agentic RAG workflows.

BERT Distributed Training Slurm Mixed Precision Qdrant

Senior Machine Learning Engineer

Jukshio Tech

Aug 2020 - Dec 2023 (3 years 4 months)

Led MLOps and engineering for a 60+ cross-functional team, established Kubeflow CI/CD and monitoring that cut operational toil 40%, deployed CV pipelines handling 4.5M+ daily transactions with 95% precision and optimized edge inference latency by 15%.

MLOps Kubeflow Computer Vision Model Quantization ETL Onnx PyTorch

Education

Degrees, certifications, and relevant coursework

University of South Dakota

Master of Science, Computer Science with Artificial Intelligence

Master of Science in Computer Science with a focus on Artificial Intelligence completed in December 2025, including research on multimodal medical report generation and applied geospatial ML.