Senior ML Engineer - Business Impact & Deployments

Maheswar Mekala Machine Learning Engineer

OH, USA +1-440-***-**** ***********@*****.*** LinkedIn

SUMMARY

Machine Learning Engineer with 4+ years of experience driving measurable business outcomes through advanced analytics and model-driven decision making. Improved purchase conversions by 17%, reduced bounce rates by 14%, and increased forecast accuracy by 23%. Experienced in building solutions that accelerate adoption, enhance transparency, and support strategic growth. Recognized for translating complex data into actionable insights, strengthening collaboration across teams, and consistently delivering high-value results that align with organizational objectives.

TECHNICAL SKILLS

Programming & Data: Python (Pandas, NumPy, Scikit-learn), SQL, Git/GitHub, Data Cleaning & Preprocessing

Machine Learning: Supervised Learning (Regression, Classification), Unsupervised Learning (K-Means, DBSCAN), Model Evaluation (Accuracy, Precision, Recall, F1, ROC-AUC)

Deep Learning: PyTorch, TensorFlow, CNNs (Image Processing), RNN/LSTM (Time-Series Modeling), Transformer Architectures

NLP & Generative AI: Hugging Face Transformers, BERT, GPT-based Models, Large Language Models (LLMs), Text Classification, Embeddings

Deployment & MLOps: Flask, FastAPI (REST APIs), Docker, MLflow, Weights & Biases, Airflow, Prefect, AWS SageMaker, EC2, S3

Cloud & Edge Deployment: AWS, GCP, Azure (Foundational), TensorFlow Lite, ONNX, Serverless Architectures (AWS Lambda)

Data Visualization & Reporting: Matplotlib, Seaborn, Power BI, Tableau

Mathematics & Statistics: Probability, Linear Algebra, Statistical Inference, Optimization (Gradient Descent), Hypothesis Testing

AI Explainability & Ethics: SHAP, LIME, Bias Detection, Model Interpretability

PROFESSIONAL EXPERIENCE

Shopify USA Jan 2025 - Current

ML Engineer

Heightened purchase conversions 17% by deploying recommendation models on AWS SageMaker integrated into live customer personalization systems.

Reduced bounce rates 14% by applying Transformer-based NLP models (BERT, GPT) to improve search accuracy and discovery relevance.

Decreased deployment cycle time 40% through end-to-end automation of ML pipelines using Airflow, MLflow, and Docker orchestration.

Enhanced transparency of deployed models by integrating SHAP and LIME explainability, strengthening trust and compliance across regulated environments.

Improved real-time inference speed 28% by optimizing scalable APIs, enabling faster predictions during peak traffic across ecommerce platforms.

Delivered measurable 11% uplift in user engagement by designing and validating A/B tests for search and recommendation engines.

Maintained predictive performance above 90% accuracy by monitoring drift and retraining models using automated MLflow pipelines.

Accelerated ML adoption by integrating engineered models directly into production applications used daily by cross-functional business stakeholders.

Developed and trained deep learning models in PyTorch to support large-scale recommendation systems.

Implemented and optimized TensorFlow models for NLP and personalization workloads deployed in production.

Inavan India Technologies, India May 2020 - Jul 2023

ML Engineer

Delivered 92%+ model accuracy by developing regression, classification, and clustering solutions for large-scale healthcare and retail datasets.

Amplified forecast precision 23% by designing CNN-based image analytics and LSTM time-series models for enterprise supply chain optimization.

Reduced latency 35% by containerizing ML APIs with Flask, FastAPI, and Docker for scalable production implementations.

Streamlined workflows by automating preprocessing, training, and retraining tasks with Airflow and Prefect, reducing manual intervention significantly.

Improved executive decision-making by building Tableau and Power BI dashboards translating predictive outputs into actionable KPIs.

Boosted model robustness by engineering domain-specific features across structured and unstructured datasets for diverse client projects.

Validated predictive reliability by applying metrics such as precision, recall, F1, and ROC-AUC across all deployed models.

Strengthened client outcomes by delivering complete ML solutions collaboratively with engineers, analysts, and business stakeholders across industries.

Built and trained deep learning models using PyTorch for image, time-series, and structured data applications.

Developed TensorFlow pipelines for model training, validation, and deployment across client ML solutions.

EDUCATION

Master's degree, Computer Science Aug 2023 - May 2025

University of Dayton (OH, USA)

Bachelor of Technology in Computer Science Jun 2019 - Jun 2023

Dayananda Sagar Institutions (Karnataka, India)

PROJECTS

Build GPT from Scratch Jul 2025 - Aug 2025

Implemented Transformer components in PyTorch, including attention and positional encoding, enabling character- and token-level language model development.

Improved training efficiency and accuracy by applying mini-batching, cross-entropy loss, backpropagation, and mixed-precision optimization during model training.

Twitter Sentiment Analysis Model Jan 2025 - Feb 2025

Built and deployed an end-to-end sentiment analysis pipeline, covering data preprocessing, model training, and web-based application integration.

Escalated classification accuracy to 80% through hyperparameter tuning and optimized preprocessing methods including tokenization and vectorization.

Exploratory Data Analysis (EDA) on Diabetes Dataset Nov 2024 - Dec 2024

Performed feature engineering and statistical analysis on diabetes dataset, uncovering trends that informed predictive model development opportunities.

Designed an interactive dashboard in Streamlit, allowing users to explore key correlations and prevalence across risk factors.

Contact this candidate