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

I am Vineet Madireddy, an AI/ML Engineer with 3+ years of experience delivering machine learning and generative AI solutions for marketing analytics and enterprise platforms. I specialize in NLP, transformer models, retrieval-augmented generation, and batch inference, with hands-on experience deploying models on AWS SageMaker, Docker, Kubernetes, and Terraform, collaborating with product and analytics teams to drive insights.…I am Vineet Madireddy, an AI/ML Engineer with 3+ years of experience delivering machine learning and generative AI solutions for marketing analytics and enterprise platforms. I specialize in NLP, transformer models, retrieval-augmented generation, and batch inference, with hands-on experience deploying models on AWS SageMaker, Docker, Kubernetes, and Terraform, collaborating with product and analytics teams to drive insights.

Vineet Madireddy

Data Scientist, AI Engineer, AI Developer, +2





I am Vineet Madireddy, an AI/ML Engineer with 3+ years of experience delivering machine learning and generative AI solutions for marketing analytics and enterprise platforms. I specialize in NLP, transformer models, retrieval-augmented generation, and batch inference, with hands-on experience deploying models on AWS SageMaker, Docker, Kubernetes, and Terraform, collaborating with product and analytics teams to drive insights.…I am Vineet Madireddy, an AI/ML Engineer with 3+ years of experience delivering machine learning and generative AI solutions for marketing analytics and enterprise platforms. I specialize in NLP, transformer models, retrieval-augmented generation, and batch inference, with hands-on experience deploying models on AWS SageMaker, Docker, Kubernetes, and Terraform, collaborating with product and analytics teams to drive insights.

Available to hire

I am Vineet Madireddy, an AI/ML Engineer with 3+ years of experience delivering machine learning and generative AI solutions for marketing analytics and enterprise platforms.

I specialize in NLP, transformer models, retrieval-augmented generation, and batch inference, with hands-on experience deploying models on AWS SageMaker, Docker, Kubernetes, and Terraform, collaborating with product and analytics teams to drive insights.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI/ML Engineer at Epsilon

October 1, 2025 - Present

Designed NLP pipelines for marketing content classification using Hugging Face Transformers and BERT, processing 420K+ customer text records per weekly campaign cycle. Trained abstractive summarization models with TensorFlow on AWS SageMaker, delivering 9 recurring weekly insight briefs for campaign planning workflows. Implemented prompt engineering and retrieval-augmented generation with GPT-based models, LangChain, and REST APIs, coordinating with product and analytics teams to handle 30+ ad-hoc insight requests per month. Adapted open-source LLMs for offline experimentation using LLaMA, performing tokenization analysis and embedding comparisons across 18 model variants. Developed transformer-based language understanding models using PyTorch, applying attention analysis and sequence-length tuning to support daily batch inference of 75K+ customer interaction summaries. Deployed batch inference outputs and model artifacts to AWS S3 and documented model behavior to support 20+ analysts

ML Engineer at Robosoft Technologies

August 1, 2020 - July 1, 2023

Trained classification models using scikit-learn with feature selection, tuning, and cross-validation for 5 customer behavior use cases. Built deep learning models with TensorFlow to process 3.6 million labeled interactions in mobile analytics. Applied gradient boosting with XGBoost to build ranking and propensity models consumed by 4 internal pipelines supporting content and notification decisions. Conducted offline model evaluation using cross-validation across 6 candidate variants and documented selection rationale. Packaged trained models for batch execution using joblib serialization with versioning for 2 deployment environments. Validated predictions through backtesting across 28 historical windows and implemented MLOps monitoring to ensure stability before release.