Rose Sharma
AI / ML Engineer
Specializing in LLMs, RAG pipelines & NLP. Turning complex AI research into production-ready, deployed applications with measurable outcomes.
AI engineer with a builder's mindset
I don't just train models — I build complete AI systems that work in the real world.
I'm a 2025 B.Tech CSE (AI Specialization) graduate from Rungta College of Engineering & Technology, Bhilai. Across 3 internships I've delivered measurable outcomes: sub-2s RAG pipelines, 90%+ accuracy NLP classifiers, and modular LLM data pipelines with robust retry logic.
I work across the full AI stack — from data preprocessing and model training to API development and cloud deployment. Every project I build has a live demo.
Currently seeking a Junior AI/ML or GenAI Engineer role. Open to relocation anywhere in India.
What I've built
Production-grade AI systems with measurable, real-world outcomes. Every project is live and deployed.
MIRA — Medical Intelligence Robotic Automation
AI-powered patient health prediction platform. Generates structured clinical reports with risk evaluation, gauge indicators, and recommendations from blood test data.
LLM Retail Assistant
Production-grade RAG pipeline for natural language product queries with multi-turn conversational AI. LangChain memory chains with Groq API for enhanced recommendation accuracy.
Fake Domain Detector
Binary ML classifier for phishing domain detection. Curated and labeled 1,000+ domain samples. Benchmarked Random Forest, SVM, Logistic Regression, and Gradient Boosting.
LLM Data Pipeline
Modular ingestion pipeline processing .txt/.pdf files and live URLs via direct LLM API calls. Tenacity-based retry logic with exponential backoff for production reliability.
Where I've worked
3 internships delivering real, production AI systems with measurable outcomes.
- Engineered Credit Scoring ML model using XGBoost and scikit-learn for financial risk classification across customer segments
- Designed Customer Churn Prediction system using Logistic Regression, Random Forest, XGBoost — 82.6% accuracy, AUC 0.795 on telecom dataset
- Trained Heart Disease classifier using 4 ML algorithms evaluated via Accuracy, Precision, Recall, F1-Score, and ROC-AUC
- Built Twitter Sentiment Analysis tool using TextBlob, Naive Bayes, and SVM with TF-IDF vectorization on 3-class NLP classification
- Developed end-to-end Customer Churn pipeline covering preprocessing, model training, hyperparameter tuning, evaluation, and visualization
- Built House Price Prediction ML pipeline using Linear Regression on California Housing dataset — R² score of 0.60 across 20,640 samples
- Deployed model via Streamlit Cloud for browser-based access with interactive user interface
Certifications
Let's work together
Actively looking for Junior AI/ML Engineer, GenAI Developer, or NLP Engineer roles.
Immediate joiner · Open to relocation anywhere in India.