AI/ML Engineer with Cloud & MLOps Expertise

Mohana Dhara

+1-980-***-**** **************@*****.*** LinkedIn GitHub

SUMMARY

AI/ML Engineer with 4 years of experience in designing, developing, and deploying machine learning models and AI driven solutions. Proficient in Python, TensorFlow, PyTorch, Scikit-learn, and OpenCV, with expertise in deep learning, NLP, computer vision, and MLOps. Skilled in end-to-end ML pipelines, including data preprocessing, model training, optimization, and deployment using Docker, Kubernetes, and CI/CD. Experienced in cloud platforms (AWS, GCP, Azure), big data technologies (Spark, Hadoop), and databases

(SQL, NoSQL, MongoDB).

SKILLS

• Methodologies and IDEs: SDLC, Agile, Waterfall, Visual Studio Code, PyCharm, Colab

• Language and Databases: Python, Go, SQL, R, C/C++, MATLAB, MySQL, Oracle, MongoDB

• Library/ Packages: PyTorch, TensorFlow, Keras, NumPy, SciPy, Pandas, Regex, SciKit-Learn, XGBoost, OpenCV, NLTK, SpaCy, Matplotlib, Ggplot, Seaborn, ResNet 50

• Tools: Power BI, Tableau, Microsoft Excel, Spark, Airflow, Snowflake, Hadoop, MLflow, Docker, Kubernetes, Google Analytics, Git, GitHub, Jira, Jenkins, ETL

• ML Algorithms: Regression, Supervised Learning, Unsupervised Learning, Random Forest, Linear Regression, Decision Tree, Deep Learning, Clustering, Classification, Time-Series, Tensorflow, Keras, NLP, GANs, Open AI, LLM, RNN, CNN

• Other Skills: Data Cleaning, Data Wrangling, Data Warehousing, Data Visualization, Critical Thinking, Communication Skills, Presentation Skills, Problem Solving, Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform, A/B Testing PROFESSIONAL EXPERIENCE

Wells Fargo Feb 2025 - Dec 2025

Machine Learning Engineer Remote, USA

• Architected a scalable AI/ML platform on AWS SageMaker, Lambda, EC2, and S3, enabling credit risk modeling, compliance automation, and personalized financial advisory using supervised, deep, and generative learning.

• Designed high-performing credit default models using XGBoost and LightGBM, improving AUC-ROC by 28% and enhancing risk stratification accuracy for loan underwriting across diverse customer segments.

• Developed explainable ML pipelines with SHAP for regulatory compliance (FCRA, Basel III, and GDPR), enabling transparent decision-making and audit-ready model interpretability for financial institutions.

• Deployed custom Transformer-based generative AI solutions (BERT variants, LangChain, OpenAI GPT, HuggingFace on AWS) for behavioral analytics and financial recommendations, achieving 30% higher user engagement, 25% better recommendation precision, and 40% faster decision cycles.

• Engineered advanced financial features including behavioral transaction embeddings and credit utilization metrics; applied PCA and UMAP to improve generalizability and computational efficiency. PWC May 2022 - Dec 2023

Data Scientist Mumbai, India

• Managed the end-to-end analysis of large-scale policyholder datasets using advanced SQL and Python (Pandas, NumPy), performing robust data wrangling, feature engineering, and EDA to uncover key drivers of customer renewal behavior.

• Designed and deployed scalable machine learning models (Logistic Regression, Random Forest, XGBoost) using Scikit-learn to forecast churn and segment customers, enabling a 20% uplift in targeted retention strategies.

• Conducted rigorous A/B testing and statistical inference (hypothesis testing, multivariate regression) using Statsmodels and SciPy to evaluate dynamic pricing and personalized renewal offers, resulting in a 15% increase in renewal rates.

• Delivered executive-level insights by building interactive dashboards in Tableau and Power BI, visualizing churn risk, policy lapse trends, and renewal KPIs to drive data-informed strategic decisions.

• Partnered with data engineering teams to automate ETL workflows using Python and SQL, integrating multi-source data (CRM, policy, claims) into a centralized AWS data lake (S3, Glue, Redshift), improving data availability and reliability.

• Continuously optimized model performance through cross-validation, ROC-AUC analysis, and confusion matrix evaluations, increasing predictive accuracy and model robustness by 10%. SRM University Jan 2021 - Mar 2022

Research Intern AP, India

• Conducted applied machine learning research in computer vision and NLP, focusing on optimization techniques for model efficiency and accuracy.

• Collected, cleaned, and preprocessed large-scale image and text datasets, applying techniques such as normalization, augmentation, tokenization, and dimensionality reduction to improve training robustness.

• Built and optimized classification models with TensorFlow and PyTorch, leveraging CNNs, RNNs, and Transformers, applying feature engineering and robust evaluation, achieving 23% performance improvement.

• Published findings in Scopus-indexed journals and IEEE Xplore, demonstrating contributions to the fields of AI-driven image recognition and natural language understanding.

• Collaborated with faculty researchers to document experiments, co-author publications, and present results in academic forums, strengthening both technical research output and scholarly communication skills. PROJECTS

Retail Store Demand Forecasting Pipeline

• Engineered a cloud-based ETL pipeline to aggregate sales data from 500+ retail stores into an AWS RDS database.

• Leveraged Pandas and NumPy to clean and preprocess sales data, ensuring consistency and accuracy.

• Implemented time-series forecasting models using Python to predict product demand, reducing 30% inventory shortages.

• Deployed pipeline components using Docker and Kubernetes for scalable, fault-tolerant operations. The NLP Assistant (RAG)Ask-My-Docs: Intelligent Document Assistant

• Engineered an end-to-end Retrieval-Augmented Generation (RAG) pipeline to query unstructured PDF data, enabling users to extract insights from complex documents with 95% relevance accuracy.

• Implemented semantic search using LangChain for recursive text chunking and ChromaDB for vector storage, optimizing retrieval latency to under 2 seconds per query.

• Mitigated LLM hallucinations by enforcing strict context-window constraints, ensuring generated answers are grounded solely in the provided source material.

• Deployed the full-stack application on Streamlit Cloud, featuring a user-friendly interface that handles file ingestion and chat history management in real-time.

The Computer Vision Monitor

• Developed a lightweight Computer Vision application that tracks user posture in real-time by analyzing 33 skeletal landmarks using MediaPipe Pose.

• Designed a custom geometric algorithm using NumPy to calculate vector angles between the ear, shoulder, and vertical axis, detecting poor posture with <100ms latency.

• Optimized model performance to achieve 30+ FPS on standard CPU hardware by utilizing grayscale preprocessing and frame skipping techniques, eliminating the need for a GPU.

• Built a background-running desktop GUI using Tkinter that triggers audio-visual alerts after sustained poor posture duration reducing false positives.

EDUCATION

University of North Carolina Jan 2024 - May 2025

Master of Science, Machine learning and artificial intelligence Charlotte, USA SRM University Jul 2019 - Aug 2023

Bachelor of Engineering, Machine learning and artificial intelligence Amaravati, India CERTIFICATIONS

• AWS Certified Solutions Architect – Associate

• Microsoft Azure Fundamentals (AZ-900)

• AI Programming with Python Nano degree (AWS x Udacity)

Contact this candidate