Career
Experience
AI & Data Science
8+ Yrs
Industrial & Clinical Leadership
Enterprise RAG Gain
+40%
Query Precision at NOV Inc.
Edge Vision QA
100%
Zero-GPU Verification Accuracy
Clinical Oncology AI
60+
Papers in Nature, Lancet & IEEE
Filter Career Sector:Showing 6 positions
- Mar 2024 – PresentCurrent
Data Scientist / AI Research Engineer
National Oilwell Varco (NOV Inc.)•Houston, TX
AI & Cloud Solutions Architect — driving enterprise AI, digital transformation, and scalable ML systems
- Leading development of an AI-powered computer vision system for Current Transformer (CT) verification using dual-camera architecture, OCR, and classical ML classifiers — achieving 100% accuracy with no GPU required.
- Architected and deployed a Retrieval-Augmented Generation (RAG) Sales Order Chatbot integrating AWS Bedrock LLMs, Athena SQL generation, and Kendra knowledge retrieval, alongside real-time anomaly detection systems.
- Built MaintGPT, an AI maintenance assistant combining LLM reasoning, audio diagnostics (YAMNet), speech-to-text (Whisper), and vector similarity search, and real-time equipment telemetry (FastAPI microservices on Docker) for industrial troubleshooting.
- Automated SQL generation with LLMs, cutting manual query handling time by 80% and improving structured query accuracy by 40% through optimized vector search, chunking, and hybrid embeddings.
- Integrated AWS services (S3, Athena, Kendra, Bedrock, ECS) for scalable enterprise AI workloads, and engineered REST APIs with FastAPI and Docker for enterprise-wide deployment.
- Built Power BI dashboards with Databricks for real-time carbon footprint analytics, and document-based chatbots that automate retrieval and improve internal communication.
- Created a time-series analytics API using Gaussian Mixture Models for anomaly detection and predictive maintenance, serving real-time predictions through RESTful endpoints.
- Designed an inventory segmentation system that combines clustering algorithms with business logic, deployed as a containerized microservice on AWS.
- Implemented end-to-end ML pipelines — data ingestion, feature engineering, training and validation, deployment, and monitoring — and tested and optimized DevOps pipelines (CI/CD, GitHub Actions, Docker).
Technologies & Architecture- AWS Bedrock
- LLMs
- RAG
- LangChain
- ChromaDB
- FAISS
- FastAPI
- Docker
- PostgreSQL
- Power BI
- Databricks
- Nov 2022 – Feb 2024
Research Scientist / Investigator
Dept. of Imaging Physics, MD Anderson Cancer Center•Houston, TX
Machine Learning & AI for cancer and healthcare research
- Developed SwarmDeepSurv, integrating swarm intelligence with deep survival networks for prognostic radiomics signatures across four solid cancers (Patterns, Cell Press).
- Designed ML models for subgroup identification and treatment effect estimation in NSCLC patients receiving immunotherapy (Nature Communications, 2025).
- Applied supervised, unsupervised, and ensemble learning, CNNs, and transformers to real-time clinical data for predictive decision-making.
- Built models for clinically actionable biomarkers with feature engineering, hyperparameter tuning, and SHAP/LIME explainability.
- Integrated multi-modal datasets (radiology, pathology, genomics, EHR) for personalized cancer treatment predictions.
- Used swarm intelligence (PSO, ACO) for feature selection on high-dimensional imaging and genomic data, with Python (scikit-learn, TensorFlow, PyTorch) pipelines integrated with CI/CD.
- Built reproducible scientific software pipelines in Python for large-scale medical data processing (imaging, genomics, clinical records).
- Co-authored 15+ peer-reviewed publications, presented at international AI conferences, contributed to NIH grant proposals on AI-driven biomarker discovery, and mentored graduate researchers on ML best practices and reproducible research.
Technologies & Architecture- PyTorch
- TensorFlow
- scikit-learn
- SHAP / LIME
- Medical Imaging
- Genomics
- Nov 2021 – Oct 2022
Postdoctoral Fellow
Dept. of Imaging Physics, MD Anderson Cancer Center•Houston, TX
- Developed computational approaches for cancer diagnosis, treatment, and prevention using AI.
- Applied swarm intelligence and deep learning for imaging biomarker discovery and survival prediction.
- Built survival models using feature selection and swarm-based strategies.
- Developed deep learning methods (CNNs, ensemble models) for CT image analysis, with data-processing pipelines for high-dimensional medical imaging datasets.
- Assisted with NIH and NSF funding proposals for AI-driven biomarker research, and trained in scientific grant writing.
Technologies & Architecture- Deep Learning
- Survival Models
- Biomarkers
- NIH / NSF Proposals
- Mar 2021 – Oct 2021
Research Scientist
Universiti Teknologi PETRONAS•Malaysia
Machine Learning & AI for oil and gas
- Led AI research as Core AI Researcher on oil and gas industry projects, including predictive analytics on pipeline weldment radiography (PrAnWeld+ Phase 1).
- Developed AI-driven defect detection and classification for radiographic imaging.
- Led research in swarm intelligence and multi-objective optimization, and developed open-source optimization libraries in Python.
- Authored research proposals for national and international funding — successful grants total USD 305K.
Technologies & Architecture- Predictive Analytics
- Oil & Gas
- Defect Detection
- Radiography
- Jul 2017 – Feb 2021
Graduate Research Assistant
Center for Research in Data Science (CeRDaS), UTP•Malaysia
- Researched multi-objective metaheuristic optimization for feature selection in classification problems.
- Developed hybrid swarm intelligence algorithms (Grey Wolf Optimizer, Particle Swarm Optimization) for high-dimensional data.
- Published 10+ peer-reviewed papers during doctoral studies.
- Dec 2012 – Aug 2014
Lecturer of Computer Science
Albaydha University•Yemen
- Taught Introduction to CS, C++, Java, Software Requirements, and Networking.
- Supervised final year projects and served as course coordinator.