Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Yashkumar Prajapati, a data-driven AI/ML engineer based in Chicago who loves turning complex data into practical products. I thrive on building end-to-end ML systems, from data preprocessing and model development to scalable deployment and monitoring in production. I specialize in Python, TensorFlow, Scikit-learn, XGBoost, and cloud platforms, and I enjoy crafting LLM inference pipelines, brand-personalized content solutions, and data-driven decision support for enterprise teams.…I'm Yashkumar Prajapati, a data-driven AI/ML engineer based in Chicago who loves turning complex data into practical products. I thrive on building end-to-end ML systems, from data preprocessing and model development to scalable deployment and monitoring in production. I specialize in Python, TensorFlow, Scikit-learn, XGBoost, and cloud platforms, and I enjoy crafting LLM inference pipelines, brand-personalized content solutions, and data-driven decision support for enterprise teams.

Yashkumar Prajapati

Data Scientist, AI Engineer, Full Stack Developer, +2





I'm Yashkumar Prajapati, a data-driven AI/ML engineer based in Chicago who loves turning complex data into practical products. I thrive on building end-to-end ML systems, from data preprocessing and model development to scalable deployment and monitoring in production. I specialize in Python, TensorFlow, Scikit-learn, XGBoost, and cloud platforms, and I enjoy crafting LLM inference pipelines, brand-personalized content solutions, and data-driven decision support for enterprise teams.…I'm Yashkumar Prajapati, a data-driven AI/ML engineer based in Chicago who loves turning complex data into practical products. I thrive on building end-to-end ML systems, from data preprocessing and model development to scalable deployment and monitoring in production. I specialize in Python, TensorFlow, Scikit-learn, XGBoost, and cloud platforms, and I enjoy crafting LLM inference pipelines, brand-personalized content solutions, and data-driven decision support for enterprise teams.

Available to hire

I’m Yashkumar Prajapati, a data-driven AI/ML engineer based in Chicago who loves turning complex data into practical products. I thrive on building end-to-end ML systems, from data preprocessing and model development to scalable deployment and monitoring in production.

I specialize in Python, TensorFlow, Scikit-learn, XGBoost, and cloud platforms, and I enjoy crafting LLM inference pipelines, brand-personalized content solutions, and data-driven decision support for enterprise teams.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

ML Engineer (Contract) at Typeface AI

August 1, 2025 - Present

Led optimization of LLM inference serving and multi-cloud integration (Vertex AI and Sagemaker) to support 12,000+ daily content-generation requests, achieving 99.8% uptime. Built and evaluated text-generation pipelines with Python and TensorFlow, enabling brand-personalized marketing content. Created an automated content quality evaluation framework using Scikit-learn and NLP metrics (BLEU, ROUGE, brand compliance), processing 25,000+ samples and reducing manual QA time by 40%. Developed data preprocessing for training data from 50,000+ marketing documents across 8 clients and contributed to REST API endpoints with Flask to generate 2,000+ content variations during pilots with 99.5% uptime.

Jr. Machine Learning Engineer at Simform Solutions Pvt. Ltd.

July 1, 2022 - July 1, 2023

Developed and deployed 12+ classification models for e-commerce and fintech clients using Scikit-learn and XGBoost, achieving an average accuracy of 82% and enabling 3 production deployments. Built ETL pipelines for 500K+ records, reducing data prep time by 35% through automated null-handling and feature engineering. Implemented Flask-based REST endpoints for real-time predictions with <200ms latency and 5,000+ daily inferences. Optimized hyperparameter tuning with GridSearchCV across 30+ parameter combinations, improving F1-scores by 8-12%. Produced EDA reports (Matplotlib/Seaborn) for 10+ client presentations and tracked experiments with MLflow.