Raghu ram Vadali - Telco Customer Churn Prediction – End-to-End ML Pipeline

Recommended expert

Munich, Germany

Experience

Jul 2025 - Jul 2025

1 month

Telco Customer Churn Prediction – End-to-End ML Pipeline

Self-Initiated Project

Designed and implemented a full machine learning pipeline for churn prediction using the Telco dataset.
Applied preprocessing techniques including missing value handling, categorical encoding, feature scaling, and PCA.
Built and compared over 15 models (logistic regression, random forest, XGBoost, etc.) and evaluated them using accuracy, precision, recall, F1 score, ROC AUC, and PR AUC.
Tuned hyperparameters with GridSearchCV, achieving 80.6% accuracy with random forest and XGBoost.
Created visual reports (bar plots, heatmaps, radar charts) to interpret model performance and churn drivers.
Exported reusable pipelines and trained models with joblib for deployment.

Mar 2025 - Mar 2025

1 month

Stock Price Analysis and Risk Modeling

Self-Initiated Project

Analyzed historical stock data from Yahoo Finance using adjusted closing prices.
Computed simple and exponential moving averages to identify market trends.
Evaluated performance via daily and cumulative returns.
Visualized correlation heatmaps and kernel density estimation (KDE) plots.
Assessed value at risk (VaR) at 95% and 99% confidence using variance-covariance, historical simulation, and Monte Carlo simulation with over 10,000 trials.
Calculated Sharpe ratio and volatility metrics to evaluate risk-adjusted returns.

Jan 2025 - Dec 2025

1 year

Food101 Multiclass Image Classification – Progressive Learning with TensorFlow (EfficientNetV2S)

Self-Initiated Project

Implemented multiclass image classification on Food101 dataset using a progressive learning strategy.
Prepared progressive datasets (10%, 50%, 100%) to enable staged model training using image_dataset_from_directory.
Built an input pipeline with 384×384 image size, batch loading, and data augmentation.
Applied EfficientNetV2S pretrained on ImageNet with hybrid pooling (GlobalAverage + Max pooling) and dropout before output.
Phase 1 (10% data): Feature extraction with frozen layers achieving 52% accuracy (15% validation).
Phase 2 (50% data): Fine-tuned last 30 layers with label smoothing achieving 80% accuracy (25% validation), improved to 83.5% with extended epochs.
Phase 3 (100% data): Fine-tuned last 30 layers with learning rate of 1e-5 achieving 83.5% accuracy (25% validation).
Achieved top-1 accuracy of 83.15% and top-5 accuracy of 96.59%.
Computed precision, recall, F1-score, and support per class.
Plotted per-class F1 score distribution and heatmaps of precision, recall, and F1.
Analyzed precision versus recall trade-offs and generated a correlation matrix of evaluation metrics.
Performed misclassification analysis to identify classes with highest errors, including top 10 misclassified categories and confusion patterns.

Feb 2015 - Present

11 years

Munich, Germany

Simulation Engineer / Project Leader

ARRK Engineering

Led crash simulations and airbag performance studies; developed predictive models with LS-Dyna and Pam-Crash.
Delivered engineering insights and recommendations to cross-functional teams and clients.

Oct 2012 - Feb 2015

2 years 5 months

Munich, Germany

CAE Specialist / Project Leader

Tecosim GmbH

Built and validated crash models.
Applied statistical methods and sensitivity studies to optimize safety performance.

Sep 2008 - Aug 2010

2 years

Chennai, India

Senior Engineer (Crash Analyst)

Renault-Nissan

Analyzed full-vehicle crash simulations.
Conducted parametric studies to improve crashworthiness.

Summary

Mechanical Engineer with 14+ years of experience in Finite Element Analysis (FEA), simulation modeling, and predictive analytics, now transitioning into AI, Machine Learning, and Deep Learning. Skilled in Python, TensorFlow, and Scikit-learn to design and train neural networks (ANN, CNN, RNN/LSTM) for regression, classification, image recognition, and time series forecasting. Completed self-initiated projects in computer vision (Food101 classification, EfficientNetV2S transfer learning), customer churn prediction, and stock risk modeling, showcasing expertise in end-to-end ML pipelines, optimization techniques, and model evaluation.

Skills

Programming Languages And Data: Python, C, Mysql.
Development Tools: Jupyter Notebook, Git, Pycharm, Vs Code.
Data Analysis And Visualization: Pandas, Numpy, Scikit-learn, Matplotlib, Seaborn.
Machine Learning: Regression, Classification, Clustering, Model Evaluation & Validation, Feature Engineering, Xgboost.
Deep Learning (Tensorflow/keras) - Ann & Cnn: Regression, Classification, Image Classification, Feature Extraction, Transfer Learning: Pretrained Models (Efficientnetv2s, Imagenet).
Deep Learning (Tensorflow/keras) - Optimization: Dropout, Weight Decay, Lr Scheduling, Label Smoothing.
Deep Learning (Tensorflow/keras) - Sequence & Time Series: Rnn, Lstm, Gru, Sliding Window Forecasting.
Deep Learning (Tensorflow/keras) - Model Evaluation: Precision, Recall, F1, Roc, Misclassification Analysis.
Mechanical Engineering: Cad, Fea (Ls-dyna, Pam-crash, Abaqus), Matlab.

Languages

Hindi

Native

Telugu

Native

English

Advanced

German

Intermediate

Education

Oct 2010 - Sep 2012

University of Stuttgart

Master of Science, Computational Mechanics of Materials and Structures · Computational Mechanics of Materials and Structures · Stuttgart, Germany

Oct 2000 - Jun 2004

Jawaharlal Nehru Technical University

Bachelor of Engineering, Mechanical Engineering · Mechanical Engineering · India

Certifications & licenses

Data Science Bootcamp

Udemy

Machine Learning A-Z: AI, Python

Udemy

Python For Data Analysis And Visualization

Udemy

TensorFlow For Deep Learning Bootcamp

Udemy

MySQL For Data Analytics & BI

Udemy

Need a freelancer? Find your match in seconds.

Try FRATCH GPT

Similar Freelancers

Discover other experts with similar qualifications and experience

Experience

Telco Customer Churn Prediction – End-to-End ML Pipeline

Self-Initiated Project

Stock Price Analysis and Risk Modeling

Self-Initiated Project

Food101 Multiclass Image Classification – Progressive Learning with TensorFlow (EfficientNetV2S)

Self-Initiated Project

Simulation Engineer / Project Leader

ARRK Engineering

CAE Specialist / Project Leader

Tecosim GmbH

Senior Engineer (Crash Analyst)

Renault-Nissan

Summary

Skills

Languages

Education

University of Stuttgart

Master of Science, Computational Mechanics of Materials and Structures · Computational Mechanics of Materials and Structures · Stuttgart, Germany

Jawaharlal Nehru Technical University

Bachelor of Engineering, Mechanical Engineering · Mechanical Engineering · India

Certifications & licenses

Data Science Bootcamp

Udemy

Machine Learning A-Z: AI, Python

Udemy

Python For Data Analysis And Visualization

Udemy

TensorFlow For Deep Learning Bootcamp

Udemy

MySQL For Data Analytics & BI

Udemy

Similar Freelancers

Projects

AI-Backend Developer Intern

Applied NLP: Word-Level Encoding for Smarter Event Predictions

Freelance Data Scientist

Principal (Data Scientist/Data Engineer/Gen AI Engineer)

Senior Data Scientist

Master's Thesis Student

GenAI Engineer

Agentic AI for a DeepResearch project

IT Project Manager

Master thesis - LLM powered RAG System

Interim Product Owner (AI Tech-Stack)

Data Partner - Computer Science - Digital Media

Master's Thesis: Analyzing Prompt Engineering for Data Extraction from Unstructured Data

Data-Scientist/AI Engineer

AI Product Manager and Engineer

Research Team Member

Forensic Accident Reconstruction Using Virtual Reality

Data Scientist Intern

Team Lead – Engagement & Relevance

Machine Learning Engineer (Part Time)

Senior MLOps, DevOps Engineer

AI Engineer (Freelance)

Implementation of Data Management Tool for LLM & Speech Technologies

Cloud (AWS) | AI | DevOps | Data

Data Scientist

AI Software Developer

Data Scientist & Data Engineer

Freelance Mathematics Expert for AI Model Training

Senior Product Manager – Revenue Management