AI Resume Screening Benchmarks: See Empikalyze in Action

Original synthetic resume

Shown in full by default. Switch to the sanitized view to see the exact text the AI evaluates.

DR. ANIRUDH RAO Lead AI/ML Engineer | Medical Imaging | Computer Vision | NLP | Deep Learning Location: Bengaluru, Karnataka, India Phone: +91 90000 34567 Email: anirudh.rao.ai@example.com LinkedIn: linkedin.com/in/anirudh-rao-ai-ml ________________________________________ PROFESSIONAL SUMMARY AI/ML Engineer and technical lead with 12+ years of experience designing, developing, researching, and deploying machine learning and deep learning solutions, with extensive specialization in medical imaging, computer vision, image processing, and natural language processing. Strong technical background in developing CNN-, GAN-, RNN-, LSTM-, and Transformer-based architectures using Python, PyTorch, and TensorFlow. Experienced in building medical image classification, segmentation, reconstruction, denoising, and abnormality-detection pipelines using MRI, CT, X-ray, and other clinical imaging data. Skilled in working with large-scale, annotated datasets, designing reproducible model-training pipelines, improving model accuracy, and optimizing inference performance for production environments. Hands-on experience applying deep learning to clinical image analysis and NLP to radiology reports and unstructured clinical text. Combines applied research with software engineering practices, including model evaluation, experiment tracking, API integration, deployment, and production monitoring. Has worked closely with radiologists, clinical researchers, data engineers, product managers, and software engineering teams to translate research prototypes into maintainable AI applications. Experienced in reviewing research publications, evaluating new model architectures, mentoring AI engineers, and presenting technical results to engineering leadership and product stakeholders. TECHNICAL SKILLS Programming and ML Frameworks • Python — Advanced • PyTorch • TensorFlow • Keras • NumPy, pandas, SciPy • OpenCV • scikit-learn • MONAI • Hugging Face Transformers Computer Vision and Medical Imaging • Medical Image Processing • Image Classification and Object Detection • Image Segmentation • Image Registration and Reconstruction • Image Denoising and Enhancement • MRI, CT, X-ray, and Histopathology Images • DICOM and NIfTI Data Handling • U-Net, 3D U-Net, ResNet, EfficientNet, Vision Transformers • Image Augmentation and Synthetic Data Generation • Model Calibration and Performance Evaluation Deep Learning Architectures • Convolutional Neural Networks (CNNs) • Generative Adversarial Networks (GANs) • Recurrent Neural Networks (RNNs) • Long Short-Term Memory Networks (LSTMs) • Transformer-Based Architectures • Attention Mechanisms • Vision Transformers (ViT) • Swin Transformers • BERT and ClinicalBERT • Transfer Learning and Fine-Tuning Natural Language Processing • Clinical Text Processing • Medical Entity Recognition • Clinical Text Classification • Radiology Report Analysis • Text Embeddings and Semantic Similarity • Sequence Labelling • Transformer-Based NLP • RNN/LSTM Sequence Models • Tokenization and Text Preprocessing Model Development and Research • Model Architecture Design • Hyperparameter Optimization • Transfer Learning • Loss Function and Optimizer Selection • Training and Validation Pipeline Development • Cross-Validation and Error Analysis • Ablation Studies • Model Benchmarking • Research Paper Review and Implementation • Statistical Evaluation of Model Performance Deployment and Engineering • FastAPI and REST APIs • Docker • Git • MLflow • CI/CD Workflows • AWS and Azure ML Services • Model Serving and Inference Optimization • Batch and GPU-Based Inference • Production Monitoring • Data and Model Versioning PROFESSIONAL EXPERIENCE Lead AI/ML Engineer — Medical Imaging and Clinical AI Medora AI Health Systems Pvt. Ltd. | Bengaluru, India January 2023 – Present Medora AI Health Systems is a fictional healthcare AI organization developing medical image analysis and clinical intelligence solutions. Key Responsibilities • Lead the development of deep learning models for medical imaging applications, including image classification, segmentation, lesion detection, image enhancement, and abnormality screening. • Design and optimize CNN-based architectures, including U-Net, 3D U-Net, ResNet, and EfficientNet, for two-dimensional and three-dimensional medical image analysis. • Develop Transformer-based vision models for medical image representation learning and classification, evaluating their performance against established CNN baselines. • Apply GAN-based methods for medical image denoising, image enhancement, and controlled synthetic-data generation to support experimentation with limited training data. • Develop clinical NLP pipelines that analyse radiology reports and extract relevant entities, observations, anatomical references, and findings from unstructured text. • Fine-tune BERT-family and other Transformer-based language models for clinical text classification and information extraction. • Maintain and optimize PyTorch and TensorFlow training pipelines, improving reproducibility, GPU utilization, training efficiency, and inference performance. • Coordinate data preparation, annotation review, augmentation, dataset partitioning, and quality checks for large medical imaging datasets. • Collaborate with clinical researchers and radiology specialists to review model errors, refine annotation requirements, and identify clinically relevant evaluation criteria. • Implement model validation workflows covering sensitivity, specificity, precision, recall, F1-score, Dice coefficient, intersection over union, and area under the ROC curve, depending on the application. • Work with software engineers to expose trained models through REST APIs and integrate inference services into larger healthcare applications. • Investigate production issues related to image formats, preprocessing discrepancies, prediction latency, model drift, and differences between training and deployment environments. • Review recent research in computer vision, medical image segmentation, GANs, and multimodal learning to identify improvements relevant to ongoing projects. • Mentor a team of six ML engineers and collaborate with engineering leadership on technical design reviews, model release criteria, project estimation, and delivery planning. Selected Achievements • Improved segmentation performance by approximately 5 percentage points in Dice score on a selected internal evaluation dataset through architecture tuning, augmentation, and revised loss functions. • Reduced average inference latency by approximately 32% for a deployed image-analysis workflow through model optimization, batching, and preprocessing improvements. • Developed reusable training and evaluation components that reduced duplicated implementation effort across multiple medical imaging projects. • Established consistent experiment tracking and model comparison practices, improving reproducibility across research and engineering teams. • Helped translate experimental deep learning models into API-based inference services used in internal clinical application workflows. Technologies: Python, PyTorch, TensorFlow, MONAI, OpenCV, NumPy, pandas, Hugging Face Transformers, FastAPI, Docker, MLflow, AWS. Senior Machine Learning Engineer — Medical Imaging and NLP VedaPoint Digital Health Technologies | Hyderabad, India July 2019 – December 2022 VedaPoint Digital Health Technologies is a fictional health technology organization working on medical imaging analytics and clinical information extraction. Key Responsibilities • Designed and trained CNN-based classification and segmentation models for CT, MRI, and X-ray image-analysis applications. • Developed medical image preprocessing workflows involving image normalization, resizing, denoising, augmentation, and conversion between clinical image representations. • Used PyTorch and TensorFlow to implement and compare deep learning architectures under different training configurations. • Implemented U-Net and related CNN architectures for anatomical segmentation and region-of-interest analysis. • Experimented with GAN-based image enhancement and synthetic image generation to assess potential improvements in model robustness. • Developed RNN- and LSTM-based sequence models for clinical text classification and structured information extraction from radiology reports. • Introduced Transformer-based NLP models for selected text-classification and clinical entity-extraction tasks. • Prepared large training and validation datasets by consolidating annotated imaging samples, checking label consistency, and preventing patient-level data leakage across splits. • Conducted hyperparameter tuning, cross-validation, error analysis, and model benchmarking. • Collaborated with data engineers to integrate feature preparation and prediction pipelines into internal applications. • Worked with domain specialists to review false-positive and false-negative predictions and refine model evaluation strategies. • Documented experiments, model configurations, data assumptions, limitations, and evaluation results for engineering reviews. Selected Achievements • Improved classification performance on a selected imaging benchmark through transfer learning and systematic hyperparameter optimization. • Reduced repeated model-development effort by reusing common preprocessing and evaluation components. • Developed a clinical text classification pipeline combining domain-specific preprocessing with deep learning-based language representations. • Improved model evaluation consistency by standardizing training, validation, and test-data handling practices. Technologies: Python, PyTorch, TensorFlow, Keras, OpenCV, CNNs, GANs, RNNs, LSTMs, Transformers, BERT, scikit-learn, Docker. Machine Learning Engineer — Computer Vision and NLP IrisArc Research Technologies | Pune, India January 2017 – June 2019 IrisArc Research Technologies is a fictional applied research and software development organization focused on computer vision and intelligent information extraction. Key Responsibilities • Developed Python-based machine learning and deep learning solutions for image classification, object detection, visual inspection, and image processing. • Implemented CNN architectures for image feature extraction and classification using TensorFlow and Keras, with later experimentation using PyTorch. • Built preprocessing workflows using OpenCV and NumPy for image normalization, geometric transformations, denoising, and augmentation. • Contributed to medical imaging proof-of-concept projects involving image classification and segmentation using annotated clinical datasets. • Experimented with GAN architectures for synthetic image generation and image-to-image transformation tasks. • Developed RNN- and LSTM-based models for sequence classification and text analysis. • Applied tokenization, feature engineering, embedding methods, and neural text classification to NLP problems. • Designed training pipelines with configurable model parameters, training logs, validation metrics, and reproducible experiment settings. • Analysed model errors and collaborated with data preparation teams to investigate image-label mismatches and unbalanced datasets. • Participated in technical reviews to assess model complexity, inference requirements, and alternative algorithms. Selected Achievements • Built reusable image preprocessing and training components for computer vision experimentation. • Improved selected image-classification results through transfer learning and dataset augmentation. • Contributed to early medical image analysis prototypes and established practical experience working with clinical image data. • Developed a foundation in neural sequence modelling that later supported clinical NLP and Transformer-based model development. Technologies: Python, TensorFlow, Keras, PyTorch, OpenCV, CNNs, GANs, RNNs, LSTMs, NumPy, pandas. Software Engineer — Image Processing and Machine Learning PixelMatrix Intelligent Systems | Chennai, India July 2014 – December 2016 PixelMatrix Intelligent Systems is a fictional software engineering company developing image-processing and automated visual analysis applications. Key Responsibilities • Developed Python modules for image ingestion, preprocessing, enhancement, segmentation, and feature extraction. • Applied traditional computer vision methods alongside early deep learning techniques for image classification and visual pattern recognition. • Used OpenCV, NumPy, and SciPy to implement filtering, thresholding, contour analysis, morphological operations, and geometric transformations. • Built and evaluated early CNN-based classification experiments using TensorFlow and Keras. • Prepared image datasets by organizing samples, reviewing labels, applying augmentation, and generating training and validation splits. • Assisted senior engineers with algorithm testing, performance measurement, and integration of image-analysis modules into software applications. • Wrote reusable Python code and unit tests to improve code reliability and simplify maintenance. • Documented preprocessing assumptions, model inputs, expected outputs, and known algorithm limitations. Selected Achievements • Developed reusable image-processing components adopted in multiple internal prototypes. • Improved consistency of image preparation and evaluation through standardized preprocessing scripts. • Gained foundational experience in computer vision, image analysis, model evaluation, and production-oriented Python development. Technologies: Python, TensorFlow, Keras, OpenCV, NumPy, SciPy, Git. SELECTED TECHNICAL PROJECTS 1. Medical Image Segmentation and Abnormality Detection Objective: Develop deep learning models that identify relevant anatomical regions and suspicious image regions in medical imaging datasets. Technical Approach • Prepared and validated CT and MRI datasets with associated annotations. • Compared U-Net, 3D U-Net, CNN-based segmentation models, and Transformer-based alternatives. • Evaluated augmentation strategies and loss functions for imbalanced segmentation datasets. • Used Dice score, IoU, sensitivity, precision, and recall to measure model performance. • Performed case-level error analysis with domain specialists. • Packaged the selected model into a versioned inference service for integration testing. Outcome: Established a repeatable image-segmentation workflow and improved evaluation consistency across model experiments. 2. GAN-Based Medical Image Enhancement Objective: Investigate whether deep generative models could improve image quality and support training-data augmentation. Technical Approach • Implemented GAN-based image-to-image transformation and enhancement experiments. • Compared generated outputs with source images using image-quality metrics and expert review where applicable. • Evaluated synthetic data for potential improvements in downstream model robustness. • Documented training instability, data limitations, and failure cases. • Compared results with conventional preprocessing and enhancement methods. Outcome: Produced a repeatable experimental framework for evaluating GAN-based image enhancement and synthetic-data approaches. 3. Clinical NLP and Radiology Report Analysis Objective: Extract relevant medical information from unstructured radiology reports to support structured reporting and search applications. Technical Approach • Developed text-cleaning and tokenization workflows for clinical narrative data. • Implemented RNN- and LSTM-based baselines for clinical text classification. • Fine-tuned Transformer-based language models for classification and entity extraction. • Evaluated precision, recall, F1-score, and error categories across relevant labels. • Built an inference API for integration testing with downstream healthcare applications. Outcome: Developed reusable clinical NLP components supporting structured information extraction from unstructured reports. 4. Scalable Model Training and Deployment Pipeline Objective: Improve the repeatability and maintainability of deep learning model development. Technical Approach • Standardized dataset configurations, training parameters, and evaluation scripts. • Integrated experiment tracking and model artifact versioning. • Packaged inference applications with Docker and exposed predictions through REST APIs. • Evaluated batch inference, preprocessing overhead, and GPU utilization. • Added logging to assist debugging and production issue investigation. Outcome: Improved reproducibility and simplified integration between model-development and software-engineering workflows. EDUCATION Bachelor of Engineering in Electronics and Communication Engineering Fictional Institute of Technology, Visvesvaraya Technological University, Karnataka, India 2010 – 2014 RESEARCH AND PROFESSIONAL DEVELOPMENT • Regularly reviews peer-reviewed publications and conference papers relating to medical image analysis, computer vision, deep learning, and clinical NLP. • Has implemented and benchmarked research approaches involving CNNs, GANs, recurrent networks, and Transformer architectures. • Experienced in translating research findings into controlled experiments with documented assumptions, baseline comparisons, and evaluation metrics. • Participates in internal technical reviews and knowledge-sharing sessions covering new model architectures, deployment approaches, and evaluation practices. ADDITIONAL INFORMATION • Experience collaborating with cross-functional teams comprising AI engineers, data engineers, application developers, product teams, and clinical subject-matter experts. • Comfortable working across research experimentation, model development, software integration, testing, and production support. • Experienced in communicating model performance, technical limitations, implementation trade-offs, and project risks to engineering stakeholders. • Willing to work from Bengaluru and collaborate with distributed engineering teams. ________________________________________ TECHNICAL EXPERIENCE SUMMARY Skill Approximate experience Python 12+ years Computer Vision 10+ years Image Processing 11+ years Deep Learning 10+ years TensorFlow 9+ years PyTorch 8+ years CNN 10+ years GAN 8+ years NLP 9+ years RNN 8+ years LSTM 8+ years Transformer architectures 7+ years Medical Imaging 9+ years

Empikalyze Screening Results

Exceptionally strong match: deep medical imaging and clinical NLP expertise across all required architectures, production deployment skills, and lead-level experience; management scope largely technical.

Original synthetic resume

Shown in full by default. Switch to the sanitized view to see the exact text the AI evaluates.

KAVYA SHARMA AI/ML Engineer | Data Science | Computer Vision & NLP Location: Bengaluru, Karnataka, India Phone: +91 90000 45678 Email: kavya.sharma.ml@example.com LinkedIn: linkedin.com/in/kavya-sharma-ml PROFESSIONAL SUMMARY AI/ML Engineer and Data Science professional with 12 years of overall experience, progressing from business data operations and analytics into predictive modelling, machine learning, and AI engineering. Brings four years of direct experience developing and improving AI/ML applications, preceded by two years in data science and six years in data analytics and associate-level roles. Experienced in developing Python-based machine learning workflows, training deep learning models, preparing large datasets, and evaluating model performance. Hands on exposure to PyTorch, TensorFlow, computer vision, image processing, NLP, and Transformer-based models, with previous experience applying CNNs, RNNs, and LSTMs to classification and prediction problems. Has contributed to medical imaging proof-of-concept applications involving image classification and segmentation, while also working on document processing, text classification, forecasting, and business analytics projects. Comfortable collaborating with software developers, data engineers, and business stakeholders to integrate model outputs into applications. Holds a Data Scientist certification and has a strong foundation in statistics, feature engineering, data preparation, model evaluation, and analytical problem-solving. Experience is broader across data science and applied machine learning than that of a specialist medical imaging researcher, with comparatively limited ownership of advanced GAN architectures, complex medical image reconstruction, and research intensive deep learning development. TECHNICAL SKILLS Programming and Data Science • Python • SQL • pandas • NumPy • SciPy • scikit-learn • Matplotlib • Seaborn Machine Learning and Deep Learning • Supervised and Unsupervised Learning • Classification and Regression • Feature Engineering • Model Selection and Evaluation • Neural Networks • CNNs • RNNs • LSTMs • Transfer Learning • Transformer-Based Models • Hyperparameter Tuning Frameworks • PyTorch • TensorFlow • Keras • Hugging Face Transformers Computer Vision and Image Processing • Image Classification • Image Preprocessing and Augmentation • Object Detection — Basic to Intermediate • Image Segmentation — Applied Project Experience • OpenCV • Medical Image Classification — Limited Project Portfolio • Image Denoising — Basic Exposure • GANs — Experimental Exposure Natural Language Processing • Text Classification • Text Cleaning and Tokenization • Named Entity Recognition — Basic Experience • Word Embeddings • Sequence Modelling • Transformer-Based Text Processing • Document Information Extraction Data Engineering and Deployment • REST APIs • FastAPI • Flask • Docker • Git • MLflow — Basic Experience • Batch Inference • Model Evaluation and Monitoring • AWS — Working Knowledge • Power BI • Microsoft Excel PROFESSIONAL EXPERIENCE AI/ML Engineer AcuityBridge AI Solutions Pvt. Ltd. | Bengaluru, India October 2022 – Present AcuityBridge AI Solutions is a fictional technology company developing applied AI solutions for document intelligence, healthcare analytics, and business automation. Key Responsibilities • Develop Python-based machine learning applications for image classification, document processing, text analysis, and predictive modelling. • Train and evaluate CNN-based models for image classification tasks using TensorFlow, Keras, and PyTorch. • Develop image preprocessing pipelines using OpenCV, including resizing, normalization, augmentation, and image-quality checks. • Contribute to a medical imaging proof of concept involving classification of selected radiology image categories using labelled datasets. • Implement and evaluate a basic U-Net-based segmentation model for a limited medical image analysis use case. • Apply transfer learning using pretrained CNN architectures to reduce training requirements and improve baseline performance. • Build NLP pipelines for document classification, information extraction, and text categorization. • Fine-tune pretrained Transformer models for text classification and document understanding tasks. • Maintain existing RNN- and LSTM-based models used for selected sequential prediction and text classification workflows. • Experiment with GAN-based image augmentation in a limited research prototype, comparing generated samples with conventional augmentation approaches. • Prepare training and validation datasets, investigate class imbalance, and review data quality issues that may affect model performance. • Evaluate models using appropriate classification and segmentation metrics, including precision, recall, F1-score, confusion matrices, and Dice score where applicable. • Collaborate with backend developers to expose model predictions through REST APIs and integrate inference functionality into internal applications. • Improve Python code quality through reusable modules, logging, testing, and Git based version control. • Investigate prediction errors and performance regressions and recommend changes to preprocessing, training parameters, and model configuration. Selected Contributions • Developed reusable preprocessing and evaluation components used across several internal ML projects. • Improved performance on a selected image-classification task through transfer learning and data augmentation. • Helped integrate NLP and classification models into internal application workflows through API-based inference. • Reduced repeated manual evaluation work by standardizing experiment summaries and model-comparison reports. Technologies Python, PyTorch, TensorFlow, Keras, OpenCV, scikit-learn, Hugging Face Transformers, FastAPI, Docker, Git. Data Scientist BlueOak Analytics Technologies | Hyderabad, India October 2020 – September 2022 BlueOak Analytics Technologies is a fictional analytics company delivering predictive modelling and data science services for healthcare, retail, and customer operations. Key Responsibilities • Developed supervised machine learning models for classification, regression, customer segmentation, and forecasting. • Used Python, pandas, NumPy, and scikit-learn for data preparation, feature engineering, model training, and evaluation. • Built neural-network prototypes using TensorFlow and Keras for selected classification problems. • Applied NLP methods to customer feedback and operational text datasets, including text preprocessing, vectorization, and classification. • Evaluated pretrained Transformer models for selected text-analysis use cases. • Performed exploratory data analysis to identify missing values, outliers, class imbalance, and inconsistencies in source datasets. • Conducted cross-validation and hyperparameter tuning to compare model performance. • Prepared model performance reports for project managers and business stakeholders. • Worked with data engineers to obtain curated datasets and resolve issues affecting model training. • Supported deployment testing by preparing serialized model artifacts and inference scripts. • Documented modelling assumptions, evaluation results, and known limitations. Selected Contributions • Built reusable data preparation and model evaluation scripts for recurring data science projects. • Improved selected classification models through feature engineering and parameter optimization. • Developed a text classification prototype that helped categorize large volumes of customer feedback. • Supported the transition of selected analytical models from notebooks into repeatable Python workflows. Technologies Python, TensorFlow, Keras, scikit-learn, pandas, NumPy, SQL, NLP libraries, Jupyter Notebook. Data Analyst Meridian Business Intelligence Services | Chennai, India October 2016 – September 2020 Meridian Business Intelligence Services is a fictional business analytics organization providing reporting and data analysis support to enterprise clients. Key Responsibilities • Analysed structured business datasets using SQL and Python to identify trends, operational issues, and performance variations. • Developed automated data preparation scripts using pandas and NumPy. • Built Power BI dashboards and recurring reports for operational and management teams. • Performed exploratory data analysis and statistical analysis to support business decisions. • Developed baseline forecasting and classification models under the guidance of senior data science professionals. • Prepared datasets for predictive modelling by handling missing values, duplicate records, inconsistent formats, and categorical variables. • Compared historical trends and statistical measures to identify unusual patterns in business data. • Worked with stakeholders to define reporting requirements and interpret analytical results. • Supported data validation and reconciliation between source datasets and reporting outputs. • Documented analytical methodologies, SQL queries, and reporting definitions. • Collaborated with engineering teams to investigate data extraction and transformation issues. Selected Contributions • Automated recurring data preparation tasks, reducing manual spreadsheet processing. • Built reusable SQL queries and Python scripts for recurring analytical requirements. • Supported early predictive analytics initiatives by preparing datasets and generating baseline model features. • Improved reporting consistency through standardized calculations and validation checks. Technologies Python, SQL, pandas, NumPy, Power BI, Excel, scikit-learn, Jupyter Notebook. Associate — Data Operations and Reporting Crestline Technology Services | Bengaluru, India October 2014 – September 2016 Crestline Technology Services is a fictional IT and business services organization supporting data operations and reporting activities for enterprise clients. Key Responsibilities • Prepared and maintained structured datasets used for business reporting. • Used SQL queries to retrieve, filter, and aggregate operational data. • Performed data cleaning, duplicate identification, and validation checks before reporting. • Maintained Excel reports and operational dashboards. • Assisted senior analysts with data extraction and report preparation. • Investigated data discrepancies and coordinated corrections with operational teams. • Documented data-processing procedures and reporting requirements. • Supported recurring data updates and quality checks under established procedures. Selected Contributions • Improved consistency in recurring reporting through standardized Excel templates. • Helped identify data-entry and source-system discrepancies before reports were distributed. • Built a foundation in SQL, data preparation, documentation, and analytical problem-solving. Technologies SQL, Microsoft Excel, basic Python, reporting tools. SELECTED TECHNICAL PROJECTS 1. Medical Image Classification Prototype Objective: Evaluate deep learning models for classifying selected medical imaging samples into predefined categories. Technical Approach • Prepared labelled image datasets and applied resizing, normalization, and augmentation. • Established a baseline using pretrained CNN architectures. • Fine-tuned a CNN model using TensorFlow and PyTorch. • Compared precision, recall, F1-score, and confusion matrices across model configurations. • Investigated misclassified samples and potential class imbalance. • Documented dataset limitations and model evaluation results. Outcome: Delivered a working proof of concept for image classification. The solution remained an experimental prototype rather than a clinically validated diagnostic system. Technologies: Python, PyTorch, TensorFlow, Keras, OpenCV. 2. Medical Image Segmentation Experiment Objective: Explore segmentation methods for identifying regions of interest in a limited set of medical images. Technical Approach • Prepared image and mask pairs for model training. • Implemented a basic U-Net architecture. • Experimented with augmentation, batch size, and loss functions. • Evaluated predictions using Dice score and intersection over union. • Reviewed segmentation errors and documented limitations. Outcome: Established practical exposure to medical image segmentation and evaluation. Further work would be required for broader datasets and production-level reliability. Technologies: Python, PyTorch, OpenCV, NumPy. 3. NLP-Based Document Classification Objective: Automatically categorize incoming business documents based on their textual content. Technical Approach • Developed text cleaning, tokenization, and preprocessing workflows. • Built baseline classification models using conventional machine learning. • Compared a recurrent neural network approach with a pretrained Transformer based classifier. • Evaluated performance using precision, recall, and F1-score. • Prepared an inference script for integration testing. Outcome: Produced a reusable document classification workflow that reduced the need for manual categorization in the tested use case. Technologies: Python, PyTorch, TensorFlow, Hugging Face Transformers, scikit-learn. 4. Business Demand Forecasting Objective: Forecast operational demand using historical business data. Technical Approach • Analysed trends, seasonality, and missing observations. • Created time-based features and established statistical baselines. • Compared traditional forecasting methods with machine learning approaches. • Evaluated prediction errors across validation periods. • Prepared visual summaries for business stakeholders. Outcome: Provided a repeatable forecasting workflow for planning and operational analysis. Technologies: Python, pandas, NumPy, scikit-learn, statsmodels. EDUCATION Bachelor of Technology in Computer Science and Engineering Fictional Institute of Engineering and Technology, India 2009 – 2013 CERTIFICATIONS • Data Scientist Professional Certificate — Fictional Professional Certification Program • Machine Learning Specialization — Online Professional Training • Deep Learning with TensorFlow and Keras — Online Professional Training • Python for Data Science — Online Professional Training TECHNICAL EXPERIENCE SUMMARY Skill Python PyTorch TensorFlow Computer Vision Image Processing Medical Imaging NLP Deep Learning CNN GAN RNN LSTM Approximate experience 10+ years 4 years 6 years 4 years 5 years 2 years 5 years 6 years 6 years 2 years 4 years 4 years Transformer architectures 3 years Experience estimates reflect overlapping use across roles and projects, rather than consecutive full-time specialization in every technology. ADDITIONAL INFORMATION • Experienced in collaborating with data engineers, application developers, and business stakeholders. • Comfortable translating business problems into data science experiments and measurable model objectives. • Familiar with research-driven experimentation, model evaluation, and technical documentation. • Has experience in applied healthcare AI prototypes but has not led a comprehensive medical imaging research programme. • Open to working from Bengaluru and collaborating with cross-functional engineering teams. PROFESSIONAL PROFILE Kavya Sharma is a multidisciplinary data and AI professional whose career developed from data operations and analytics into data science and AI/ML engineering. Her strongest experience includes Python, applied machine learning, data preparation, model evaluation, NLP, and CNN-based image classification. She has relevant exposure to PyTorch, TensorFlow, medical image processing, segmentation experiments, RNN/LSTM models, Transformers, and limited GAN experimentation. However, her four years in AI/ML engineering have not been exclusively dedicated to medical imaging, and her experience in advanced generative architectures, clinical imaging research, and production-scale medical image systems is comparatively limited. Her Data Scientist certification, broad technical background, and experience across analytics and modelling make her a credible applicant for applied AI/ML positions. For this specific role, she is a moderate match because the position expects deep, sustained expertise across medical imaging, computer vision, NLP, and multiple deep learning architectures.

Empikalyze Screening Results

Strong applied AI/ML skills across CNN, NLP, Transformers; medical imaging, GAN depth, and management experience fall short of senior requirements.