Available to hire
I’m Mark Monroe Gray, an AI/ML engineer with 10+ years of hands-on experience delivering production-grade systems across financial services, healthcare, and compliance-driven environments. I design end-to-end ML pipelines—from high-dimensional data selection and feature engineering to model development, deployment, integration, and monitoring—leveraging Python, PyTorch, TensorFlow, Spark, and BigQuery on AWS and GCP.
My recent work includes transformer fine-tuning (BERT, T5, GPT), real-time scoring APIs with XGBoost and LightGBM, RAG-style retrieval systems with FAISS + LLMs, and fairness-aware pipelines with adversarial debiasing, SHAP, and SR 11-7 compliant governance. I prioritize explainability, auditability, and scalable design, working closely with product, legal, and clinical teams to turn AI into trusted, measurable outcomes.
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Language
English
Fluent
Work Experience
Senior AI Engineer at Zest AI
September 1, 2023 - PresentEngineered AWS ETL infrastructure supporting bureau and alternative data to scale inclusive credit scoring pipelines. Fine-tuned credit-domain transformers with SHAP explainability for regulatory compliance. Built RESTful microservices for ensemble inference supporting real-time decisioning. Launched fairness-first ZAML FairBoost tooling for debiasing and monitoring. Developed GenAI credit intelligence platform 'LuLu Strategy' using FAISS and GPT for policy simulation. Implemented SageMaker-based monitoring, automated retraining, and model risk documentation supporting lender audits. Mentored junior engineers on fairness and generative retrieval architectures.
ML Engineer at Lunit
August 31, 2023 - August 27, 2025Developed and fine-tuned ResNet and DenseNet classifiers for FDA-cleared diagnostic pipelines in chest X-rays. Engineered clinical-scale DICOM ingestion pipelines standardizing multi-vendor imaging data. Integrated Grad-CAM explainability for clinical diagnostics and regulatory transparency. Managed ML pipelines on Google Cloud Vertex AI and MLflow for radiology AI workflows. Led R&D NLP pilots using credit-tuned transformers for EHR structuring. Coordinated multi-site clinical studies and partnered with radiologists for labeling consistency and global hardware integrations.
Data Scientist (R&D) at Aidoc
July 31, 2019 - August 27, 2025Researched and deployed deep learning pipelines for radiology anomaly detection including ICH and PE. Engineered LSTM autoencoders for vital sign time-series anomaly detection. Developed transfer learning CNN models and synthetic data pipelines with GANs and VAEs to improve class balance. Delivered computer vision solutions for radiology quality checks, enhancing defect detection. Contributed to AWS-integrated enterprise AI platform streamlining clinical workflows. Utilized SHAP explainability with interactive dashboards. Collaborated on multi-site pilot studies for FDA submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 30, 2016 - August 27, 2025Selected neural network architectures and developed baseline models for AI tasks. Built RESTful microservices for model serving and integrated with cloud platforms. Designed large-scale ETL pipelines connecting HDFS, Hive, and databases. Automated visualizations for model metrics using Python libraries. Engineered statistical and probabilistic prediction workflows with rigorous validation. Developed interactive dashboards using JavaScript and D3.js. Implemented containerized deployment for inference services and developed ML workflow CI/CD pipelines to support DevOps transition.
Senior AI Engineer at Zest AI
September 1, 2023 - PresentLead engineering of ETL infrastructure on AWS integrating bureau and alternative data to scale inclusive credit scoring pipelines for thin-file segments. Fine-tuned credit-domain transformers (BERT/T5) in PyTorch and TensorFlow performing SHAP-based explanations aligned with ECOA/FCRA. Developed RESTful microservices for hybrid XGBoost + LightGBM inference supporting real-time decisioning including bureau and Plaid integrations. Launched ZAML FairBoost, enforcing adversarial debiasing and automated disparate impact tests. Integrated with Temenos Loan Origination to embed credit and fraud decisioning, achieving 60–80% automation with reduced charge-offs. Architected parity-aware features incorporating fairness constraints, implemented production monitoring with SageMaker Model Monitor & Clarify for drift detection, and automated SR 11-7 compliant model risk documentation generating explainability dashboards. Collaborated on fairness governance playbooks with product, legal, and compli
ML Engineer at Lunit
August 31, 2023 - August 27, 2025Developed and fine-tuned ResNet/DenseNet classifier pipelines for FDA 510(k)-cleared diagnostic tools focused on chest X-rays detecting emergent findings. Led design and testing of deep CNN models with attention and Grad-CAM explainability integrated into PACS for hospital workflows and FDA validation. Engineered clinical-scale DICOM ingestion pipelines standardizing multi-vendor scanner data for training and augmentation. Managed ML pipelines on Google Cloud Vertex AI and MLflow for scheduling retraining and version tracking. Containerized inference endpoints using Docker for scalable clinical integration. Partnered with radiologists for annotation protocol development ensuring dataset consistency for regulatory submissions. Supported global deployments and integrations with hospital-grade hardware, including Samsung X-ray devices for ER/ICU triage.
Data Scientist (R&D) at Aidoc
July 31, 2019 - August 27, 2025Researched and deployed deep learning pipelines for radiology anomaly detection including intracranial hemorrhage and pulmonary embolism. Engineered LSTM and variational RNN models for patient vital-sign time-series anomaly detection. Prototyped transfer-learning CNNs applying structured data augmentation to improve generalization. Developed seq2seq NLP models for radiology report standardization aiding documentation workflows. Created GAN and VAE-based synthetic data pipelines to mitigate class imbalance. Delivered computer vision solutions improving radiology quality control reducing manual review times. Contributed to aiOS™ platform integration with hospital PACS/RIS enhancing clinician workflows. Analyzed large datasets using R, Python, and SQL with SHAP explainability and created interactive dashboards. Supported multi-site pilot validations ahead of FDA submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 30, 2016 - August 27, 2025Developed neural network architectures (MLPs, CNNs) fundamental to AI toolkits and baseline models using Theano and scikit-learn. Created RESTful model-serving microservices integrated into Pivotal Cloud Foundry for scalable deployment. Designed data ingestion connectors for HDFS, Hive, MySQL, and Oracle underpinning large-scale ETL and feature extraction workflows. Automated analytics visualizations and statistical reports leveraging Python and R for model evaluation. Engineered probabilistic and statistical model selection tools and implemented MCMC simulations supporting fraud detection engines. Developed interactive dashboards using JavaScript and D3.js for monitoring model performance. Led Docker container deployments on GPU workstations for reproducible inference testing and resilient production deployment. Architected Jupyter Notebook integration offering secure multi-user exploratory data science workflows. Built and maintained CI/CD pipelines to support robust DevOps practices
Senior Data Scientist at Zest AI
September 1, 2023 - PresentEngineered AWS-based ETL for bureau and alternative data; scaled credit-scoring pipelines; fine-tuned credit-domain transformers (BERT/T5) for underwriting inputs with SHAP explanations aligned to ECOA/FCRA; built RESTful microservices delivering hybrid XGBoost + LightGBM inference with dense embeddings for real-time decisioning; launched ZAML FairBoost for adversarial debiasing and automated disparate impact tests; integrated with Temenos LO and Zest Protect for real-time decisions; deployed monitoring with SageMaker Model Monitor & Clarify and BigQuery pipelines for cross-cloud analytics.
ML Engineer at Lunit
August 1, 2023 - September 11, 2025Developed and fine-tuned ResNet/DenseNet-based classifiers for early triage in INSIGHT CXR; implemented Grad-CAM explainability; engineered clinical-scale DICOM ingestion pipelines; deployed on Google Cloud Vertex AI with Kubernetes; containerized inference endpoints and PACS integration; managed MLflow-based retraining workflows.
Data Scientist (R&D) at Aidoc
July 1, 2019 - September 11, 2025Researched and deployed deep learning pipelines for radiology anomaly detection (ICH, PE); engineered LSTM autoencoders for vital-sign time-series; developed transfer-learning CNNs; created GAN/VAE synthetic data pipelines to address class imbalance; delivered QA pipelines with SHAP explanations and dashboards for FDA/CE documentation and multi-site pilots.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 1, 2016 - September 11, 2025Assisted in selecting neural architectures; developed RESTful model-serving microservices; built data ingestion connectors (HDFS, Hive, MySQL, Oracle); automated visualizations of model metrics; designed feature extraction and dimensionality-reduction workflows; containerized inference services and implemented CI/CD for production deployment.
Senior Data Scientist at Zest AI
September 1, 2023 - PresentEngineered ETL infrastructure on AWS supporting bureau and alternative data, scaling inclusive credit scoring pipelines for thin-file segments. Fine-tuned credit-domain transformers (BERT/T5) in PyTorch/TensorFlow to convert textual underwriting inputs into risk signal embeddings with SHAP-based explanations aligned with ECOA/FCRA requirements. Built RESTful microservices delivering hybrid XGBoost + LightGBM ensemble inference, enriched with dense neural embeddings and SHAP explanations to support real-time bureau and Plaid integrations for sub-second decisioning. Launched ZAML FairBoost to systematically enforce adversarial and joint-optimization debiasing, and conduct automated disparate impact tests and less-discriminative alternative model searches. Implemented production-grade monitoring with SageMaker Model Monitor & Clarify, integrated with CloudWatch and Prometheus for automated retraining and rollback. Migrated selected ML pipelines to Google Cloud Platform (GCP) Vertex AI for
ML Engineer at Lunit
August 1, 2023 - September 11, 2025Developed and fine-tuned ResNet/DenseNet-based classifiers for early triage in Lunit INSIGHT CXR (FDA clearance in 2021), optimized for emergent findings such as pneumothorax and pleural effusion. Led deep CNN model design with attention modules and Grad-CAM explainability, packaging for hospital PACS integrations and FDA validation pipelines. Engineered clinical-scale DICOM ingestion pipelines, exported metadata for model training and cross-site validation, and integrated Grad-CAM overlays for regulatory transparency. Managed ML pipelines on Google Cloud Vertex AI and MLflow, orchestrated inference workloads with Kubernetes (GKE) for scalable hospital deployments, and prototyped Docker-based endpoints for PACS integration. Led concept NLP pilots to derive structured EHR outputs from radiology reports, and presented at major radiology conferences to advance clinical adoption.
Data Scientist (R&D) at Aidoc
July 1, 2019 - September 11, 2025Researched and deployed deep learning pipelines for radiology anomaly detection, including intracranial hemorrhage (ICH) and pulmonary embolism (PE). Engineered LSTM autoencoders and variational RNNs for modeling vital-sign time-series, enabling detection of physiological anomalies in hospital data streams. Prototyped transfer-learning CNNs for automated imaging workflows, and designed seq2seq NLP models to standardize radiology report phrases. Developed GAN- and VAE-based synthetic data pipelines to mitigate class imbalance, and delivered computer-vision solutions for radiology QA to reduce manual reviews. Contributed to AWS-integrated platforms for PACS/RIS integration and clinician workflows, with SHAP explainability and interactive dashboards for regulatory transparency. Led multi-site pilot studies to validate model performance and support regulatory submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 1, 2016 - September 11, 2025Assisted in selecting neural network architectures, developed baseline models, and designed RESTful model-serving microservices. Implemented data ingestion connectors for HDFS, Hive, MySQL, and Oracle to enable large-scale ETL pipelines and feature extraction. Automated diagnostics visualizations for ROC/PR and feature correlations, and built a feature-analysis engine (Pearson, covariance, PCA) within Chorus to quantify variance and generate probabilistic scoring. Created Docker-based GPU inference environments for reproducible testing and deployed CI/CD pipelines to production clusters. Built UI modules and Jupyter integrations to enable business users to interact with predictive models, supporting Alpine’s transition to DevOps-driven deployment.
Senior Data Scientist at Zest AI
September 1, 2023 - PresentLed AWS-based ETL infrastructure for bureau and alternative data; fine-tuned credit-domain transformers (BERT/T5) in PyTorch/TensorFlow to produce risk embeddings with SHAP explanations aligned to ECOA/FCRA; built RESTful microservices for hybrid XGBoost+LightGBM inference with dense embeddings and SHAP explanations for real-time bureau integrations and sub-second decisioning; launched ZAML FairBoost for adversarial debiasing and automated disparate impact testing; integrated Temenos LOS and Zest Protect for real-time decisioning; developed the LuLu Strategy module using FAISS + summarization for policy simulation and risk diagnostics; enforced fairness in feature engineering; implemented production-grade monitoring with SageMaker Model Monitor/Clarify, CloudWatch, and Prometheus with automated retraining and rollback; prototyped risk features in BigQuery ML and conducted cross-cloud benchmarking; migrated pipelines to Google Cloud Vertex AI; built BigQuery pipelines unifying bureau an
ML Engineer at Lunit
August 1, 2023 - September 11, 2025Led deep CNN classifier pipelines for INSIGHT CXR triage (FDA clearance), optimizing for emergent findings (pneumothorax, pleural effusion); implemented DICOM ingestion pipelines, metadata export for training, and Grad-CAM explainability for PACS integrations; deployed using Google Cloud Vertex AI and MLflow with Kubernetes (GKE) for scalable inference; containerized models with Docker and orchestrated inference workloads; explored NLP integration to generate structured EHR outputs from radiology reports and notes; collaborated with radiologists for labeling protocols and multi-site validation; supported global clinical deployments and hardware partnerships.
Data Scientist (R&D) at Aidoc
July 1, 2019 - September 11, 2025Researched and deployed deep learning pipelines for radiology anomaly detection (ICH, PE); engineered LSTM autoencoders for vital-sign time-series; prototyped transfer-learning CNNs for automated imaging workflows; designed seq2seq NLP models to standardize radiology reports; developed GAN/VAE synthetic data pipelines to mitigate class imbalance; delivered computer-vision solutions for quality checks; contributed to AWS-integrated aiOS platform and hospital PACS/RIS integrations; performed data mining in R/Python/SQL; used SHAP for explainability and created dashboards; partnered with radiology centers for multi-site pilots and regulatory-ready submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 1, 2016 - September 11, 2025Assisted in selecting neural nets, developed baseline models, and created RESTful model-serving microservices; integrated model services with Pivotal Cloud Foundry; built data ingestion connectors for HDFS, Hive, MySQL, and Oracle for large-scale ML ETL; automated visualization of metrics (ROC/PR, feature correlations); implemented statistically-grounded prediction pipelines with hypothesis testing and bootstrap confidence intervals; built dashboards for monitoring pipeline execution and model performance; containerized inference services with Docker for GPU hardware; integrated Jupyter into Alpine Chorus for multi-user reproducible DS workflows; established CI/CD pipelines for deploying containerized services.
Senior Data Scientist at Zest AI
September 1, 2023 - PresentEngineered ETL infrastructure on AWS supporting bureau + alternative data, scaling credit scoring pipelines for thin-file segments. Fine-tuned credit-domain transformers (BERT/T5) in PyTorch/TensorFlow to convert underwriting inputs into risk signal embeddings with SHAP-based explanations aligned with ECOA/FCRA requirements. Delivered hybrid XGBoost + LightGBM ensemble inference with dense neural embeddings and SHAP explanations for real-time bureau and Plaid integrations. Launched ZAML FairBoost to enforce adversarial and joint-optimization debiasing with automated disparate impact tests and LDA model searches. Integrated with Temenos LO for real-time credit decisioning, achieving ~60–80% automation. Built a GenAI credit intelligence module (LuLu Strategy) using FAISS vector retrieval with T5/GPT summarization for policy simulation and risk diagnostics.
ML Engineer at Lunit
August 1, 2023 - September 11, 2025Developed and fine-tuned ResNet/DenseNet-based classifiers for early triage in INSIGHT CXR (FDA clearance), optimized for emergent findings. Led model design and testing, integrated Grad-CAM explainability, and packaged for hospital PACS integrations. Built clinical-scale DICOM ingestion pipelines, exported metadata for model training, and enabled cross-site validation. Managed inference workloads on Kubernetes (GKE) and deployed Docker-based endpoints for scalable clinical integration; implemented GPU-accelerated inference with low latency on hospital hardware. Explored NLP integration for EHR outputs from radiology reports using transformer-based methods.
Data Scientist (R&D) at Aidoc
July 1, 2019 - September 11, 2025Researched and deployed deep learning pipelines for radiology anomaly detection (ICH, PE). Engineered LSTM autoencoders for patient vital sign time-series and transfer-learned CNNs for automated imaging workflows. Developed seq2seq NLP models for standardizing radiology report phrases and contributed to FDA-cleared triage tools. Built synthetic data pipelines (GAN/VAEs) to mitigate class imbalance and created hospital-facing dashboards to support regulatory submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 1, 2016 - September 11, 2025Supported model-serving microservices with Flask/Django and Chorus, enabling scalable deployment on Pivotal Cloud Foundry. Built data ingestion connectors (HDFS, Hive, MySQL, Oracle) for large-scale ETL and feature extraction. Automated model metrics visualization (ROC/PR), implemented feature analysis and cross-validation pipelines, and containerized inference services with Docker for reproducible testing. Created UI touchpoints to translate predictive insights into business decisions and built CI/CD pipelines for production deployments.
Senior Data Scientist at Zest AI
September 1, 2023 - PresentEngineered ETL infrastructure on AWS supporting bureau + alternative data (payment, utility, telecom, consented app notes), scaling inclusive credit scoring pipelines for thin-file segments. Fine-tuned credit-domain transformers (BERT/T5) in PyTorch/TensorFlow to convert underwriting inputs into risk signal embeddings with SHAP-based explanations aligned with ECOA/FCRA requirements. Built RESTful microservices delivering hybrid XGBoost + LightGBM ensembles, enriched with dense neural embeddings and SHAP feature explanations—supporting real-time bureau and Plaid integrations for sub-second decisioning. Launched ZAML FairBoost, Zest’s fairness-first tool, to systematically enforce adversarial and joint-optimization debiasing, and conduct automated disparate impact tests and less-discriminative alternative (LDA) model searches. Integrated native Temenos Loan Origination (Apr 2025), embedding in LOS credit and realtime fraud decisioning powered by Zest Protect—achieving ~60–80% aut
ML Engineer at Lunit
August 1, 2023 - September 11, 2025Led deep CNN model design and testing for early triage in Lunit INSIGHT CXR (FDA 510(k) clearance, Nov 2021), optimizing for emergent findings such as pneumothorax and pleural effusion. Engineered clinical-scale DICOM ingestion pipelines, exported metadata for model training, image augmentation, and cross-site validation. Integrated GradCAM explainability overlays into inference outputs for PACS-ready diagnostics, and managed ML pipelines on Google Cloud Vertex AI and MLflow for hyperparameter tuning and scheduled retraining. Orchestrated model inference workloads with Kubernetes (GKE) for scalable deployment across hospital PACS integrations. Prototyped Docker-based inference endpoints and containerized models for clinical integration. Built explainable AI mechanisms and optimized GPU inference with TensorRT for low-latency near real-time triage.
Data Scientist (R&D) at Aidoc
July 1, 2019 - September 11, 2025Researched and deployed deep learning pipelines for radiology anomaly detection, including intracranial hemorrhage (ICH) and pulmonary embolism (PE). Engineered LSTM autoencoders and variational RNNs for modeling patient vital sign time-series; prototyped transfer-learning CNNs for automated diagnostic imaging workflows and applied structured data augmentation to improve generalization. Designed seq2seq NLP models with attention for standardized radiology report phrasing, and contributed GAN/VAE-based synthetic data pipelines to address class imbalance. Delivered computer-vision solutions for radiology QA and contributed to AWS-integrated aiOS enterprise platform, enabling always-on triage and PACS/RIS integration. Collaborated with radiology centers on multi-site pilots to validate thresholds and support regulatory submissions.
Software Engineer (AI Tools & Visualization) at Alpine Data Labs
June 1, 2016 - September 11, 2025Assisted in selecting neural network architectures and built baseline models for exploratory analytics within Chorus. Implemented RESTful model-serving microservices using Flask/Django and containerized them for deployment on Pivotal Cloud Foundry. Designed data ingestion connectors (HDFS, Hive, MySQL, Oracle) and automated visualization pipelines to monitor ROC/PR metrics, feature correlations, and distribution patterns. Engineered prediction workflows and probabilistic scoring pipelines, developed Jupyter Notebook integration for secure multiuser scripting, and built polished CI/CD pipelines to deploy containerized inference services for production use.
Education
Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - December 31, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - December 31, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - December 31, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - December 31, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - January 1, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - January 1, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - January 1, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - January 1, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - December 31, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - December 31, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - January 1, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - January 1, 2013Master's Degree in Computer Science at The University of Texas at Austin
January 1, 2014 - January 1, 2015Bachelor's Degree in Computer Science at The Angelo State University
January 1, 2010 - January 1, 2013Qualifications
Industry Experience
Financial Services, Healthcare, Life Sciences, Software & Internet, Professional Services
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Hire a AI Engineer
We have the best ai engineer experts on Twine. Hire a ai engineer in Washington today.