Open to opportunities

Dilkhush User

@dilkhushuser1

Message

Data scientist specializing in ML, NLP, and scalable production MLOps to drive actionable business impact.

India

Message

What I'm looking for

I seek roles where I can build and deploy ML/NLP solutions in production, work with large-scale data, and grow in MLOps and LLM-driven product development.

I am a data scientist with a foundation in chemical engineering from IIT Roorkee and hands-on experience designing ML systems in production. I focus on building scalable models, deploying reliable MLOps pipelines, and turning large datasets into business value.

At GreedyGame I designed anomaly-detection systems that yielded significant cost savings and built an internal LLM-powered analytics chatbot that reduced analyst dependency substantially. I regularly work with BigQuery, PostgreSQL, GCP, and Docker to analyze multi-terabyte datasets and streamline deployments.

Previously, as a data analyst at The Smart Cube, I delivered high-accuracy NLP solutions using BERT, LSTM, Word2Vec and TF-IDF, and led projects that improved operational efficiency such as checkout optimization. As an intern, I applied ensemble models to predict shipment ETAs and processed very large datasets for modeling.

I contribute to LLM and LangChain-based customer-service automation projects and remain engaged with academic and event initiatives from my alma mater. I bring a practical, results-driven approach to ML engineering, NLP, and data-driven product improvements.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

GreedyGame

Dec 2023 - Present (2 years 2 months)

Designed and deployed anomaly detection models that prevented fraudulent transactions and achieved $340K in annual cost savings; built an LLM-powered analytics chatbot that reduced analyst dependency by 65% and improved ad targeting efficiency by 16% through user segmentation.

Anomaly Detection Isolation Forest LLM BigQuery GCP MLOps CI CD SQL Python

Data Analyst

The Smart Cube

Jun 2019 - Sep 2020 (1 year 3 months)

Implemented deep learning and statistical NLP solutions (BERT, LSTM, Word2Vec, TF-IDF) achieving 91% accuracy for product-description similarity and led a store check-out optimization that reduced manpower by 6%.

NLP BERT Word2Vec TF IDF Python Data Analysis

Data Science Intern

Roambee Corporation

Apr 2018 - Jul 2018 (3 months)

Predicted shipment ETAs using ensemble models (Random Forest Regression) achieving RMSE of 1.8 hours and preprocessed a 210M-row dataset to support model training for frequent routes.

Random Forest regression data preprocessing Python SQL