Nagur Shareef Shaik

478 Lindbergh Pl NE, Lindbergh, GA 30324 · (404) 203-9276 · shaiknagurshareef6@gmail.com · Download Resume

As a CS PhD Student and Machine Learning Specialist with 5 years of experience, I bridge the gap between cutting-edge research and production-grade systems that unlock business value across healthcare, finance, and retail. Backed by 20+ peer-reviewed publications (1000+ citations, h-index 12) at top-tier venues, my expertise spans the full ML stack, from custom neural network design to fine-tuning multimodal LLMs and architecting RAG and Agentic AI workflows. I excel at guiding models from prototype to deployment, collaborating with cross-functional teams to balance accuracy, latency, and cost at scale. Thriving where intellectual curiosity meets pragmatic engineering, I am a research-minded applied scientist who ships dependable products and effectively communicates strategy to both technical and executive audiences.

Education 🎓

Georgia State University Atlanta, GA, USA

Doctor of Philosophy (Ph.D.) in Computer Science
Expected GPA: 4.17 / 4.3
Coursework: Advanced Machine Learning, Deep Learning, Computer Vision, Natural Language Processing, Digital Image Processing, Advanced Image Processing, Computational Intelligence, Privacy Aware Computing, Fundamentals of Data Science.
Dissertation: My Ph.D. research advances multi-modal medical report generation by focusing on clinical robustness and temporal dynamics. My work involves architecting novel attention mechanisms for efficient multi-modal fusion, creating models robust to missing modalities, and training confounder-free systems to ensure causally-sound longitudinal medical reporting, accurately narrating disease progression over time.
August 2025 - December 2026

Georgia State University Atlanta, GA, USA

Master of Science in Computer Science
GPA: 4.17 / 4.3
Coursework: Advanced Machine Learning, Deep Learning, Computer Vision, Natural Language Processing, Digital Image Processing, Computational Intelligence, Fundamentals of Data Science
Thesis: Integrating multi-modal data - such as medical images, clinical text, and genetics - through scalable attention networks and vision-language models has the potential to significantly enhance diagnostic accuracy and decision-making in schizophrenia diagnosis and medical report generation. By advancing deep learning techniques for multi-modal learning and focusing on explainable AI systems, this research aims to create efficient, robust, and responsible AI solutions that improve diagnostic precision and automate medical report generation, all while ensuring deployment in resource-constrained environments.
August 2023 - May 2025

Vignan's Foundation for Science, Technology & Research Andhra Pradesh, India

Bachelors of Technology (Hons.) in Computer Science & Engineering 🏅 🏆
CGPA 9.79 / 10 (3.92 / 4.0)
Coursework: Data Structures, Design & Analysis of Algorithms, Probability & Statistics, Object Oriented Programming, Operating Systems, Database Management System, Data Mining, Computer Graphics, Artificial Intelligence, Pattern Recognition, Optimization Techniques, Computer Networks, Internet of Things
June 2016 - May 2020

Experience

Graduate Research Assistant

TReNDS Center (Georgia Tech, Georgia State & Emory) Atlanta, GA, USA
  • Causal Longitudinal Reporting: Engineered a Deep Learning framework using EBMs to generate causally-aware reports from longitudinal data while mitigating time-varying confounders.
  • Disentangled Vision Language Alignment: Developed a Multi-modal Variational Autoencoder with Mixture-of-Experts (MoE) to disentangle modality-specific features from shared features, enabling precise cross-modal alignment in radiology report generation.
August 2025 - Present

Associate III, Data Science

UST Global Inc. (Client: Elevance Health) Atlanta, GA, USA
  • Data Map Co-Pilot: Architected an Agentic AI tool using Google Agentspace to perform data profiling and map to a BigQuery data warehouse on GCP, saving 4 BSA FTEs and $250K annually.
  • RFI/RFP Response Engine: Delivered a RAG-powered engine on Azure that uses a source-linked knowledge base to generate editable, traceable, and compliance-ready responses, cutting authoring time by 60%.
  • Code Crafter: Lead a 10-person AI team to architect CodeCrafter, an Agentic AI module for UST Aplha AI platform, securing $18 million in client contracts by delivering impactful POCs.
  • Code Modernization: Architected GenAI driven code modernization solution to migrate complex legacy systems (.NET4 to .NET8; COBOL to React/Java Spring Boot) from monolithic to scalable, maintainable microservice architectures.
  • Spes2Code: Engineered a LangGraph-based Agentic Framework to generate 95% functional code by auto-scaffolding enterprise-grade NX Monorepos (React Native, Nest.js), accelerating development by 70% at 60% cost.
  • NL2SQL: Pioneered advanced data solutions by collaborating with Stanford AI Lab on a Text-to-SQL framework (benchmarked on BIRD) and developing an NL2SQL AgenticRAG POC saving 4 FTEs annually.
  • Data Clean-up: Developed SQL data clean-up script to resolve commission payment inconsistencies in health insurance policies, preventing over $500K in overpayments and streamlining processing time.
July 2024 - August 2025

Graduate Research Assistant

TReNDS Center (Georgia Tech, Georgia State & Emory) Atlanta, GA, USA
  • Medical Vision Language Transformer: Developed a novel Abstractor-Adaptor framework for resource-constrained environments, enhancing feature focus and fusion for expert-level medical image captioning accuracy.
  • Multi-Modal Imaging Genomics Transformer: Pioneered a fusion model combining genomics with sMRI and fMRI, elevating schizophrenia diagnosis accuracy by 2.12% and uncovering vital neuro-genetic markers.
  • Multi-Modal Medical Transformer: Designed a vision-language model integrating retinal image features with clinical keywords, resulting in a 13.5% increase in BLEU-4 score over GPT-2 for diagnostic report generation.
  • Guided Context Gating: Innovated an advanced attention model that optimizes context learning in retinal images, amplifying diagnosis accuracy by 2.63% over advanced attention methods and 6.53% over Vision Transformers.
  • Spatial Sequence Attention Network: Formulated a unique attention mechanism to highlight schizophrenia-specific regions in brain sMRI, increasing diagnostic accuracy by 6.52% and providing critical neuroanatomical insights.
  • Cancer Assist: Developed a ML framework for an intra-operative intelligent system aiding surgeons in cancer surgery.
August 2023 - May 2025

Software Engineer III

Carelon Global Solutions (subsidiary of Elevance Health) Hyderabad, Telangana, India
  • COmpensation INcentive System: Implemented RESTful APIs for a microservices-based application to validate, compute, and expedite incentive payments, achieving a 10% reduction in processing time and enhancing scalability.
  • Data Cleaning: Automated transactions data clean-up using Python & SQL improving operational efficiency and saving 30% of incentive over payments.
  • Resolved critical production issues, preventing $1.5M commission over payments, ensuring seamless business operations
September 2022 - August 2023

Software Engineer (Analytics & Insights)

Tata Consultancy Services Hyderabad, Telangana, India
  • Jeopardy Automator: Designed an Bug Root Cause Prediction System based on logs, by implementing Attention LSTM model in Azure ML Studio, cutting the debugging efforts and saving 3 Full-Time Equivalents annually.
  • Order Data Orchestrator: Optimized data pipelines to streamline order orchestration, reducing fallouts by 30% and ensuring seamless real-time data flows across hybrid IT environments.
  • Operational Dashboards: Achieved a $3M revenue profit increase through enhanced operational transparency and data-driven decision-making enabled by interactive dashboards visualizing order trends and business insights.
  • Auto Deployer: Implemented an Azure DevOps Model Deployment pipeline, reducing deployment time from 2 to 1.2 hours and increasing system availability by 25%.
August 2020 - September 2022

Skills

Programming Languages
  • Python, Java, C Programming

AI & Machine Learning
  • PyTorch, TensorFlow, Keras, Scikit-learn, OpenCV, NLTK, SpaCy, NumPy, Pandas, Matplotlib
  • Deep Learning (Neural Networks, Transformers, LLMs - Huggingface), Computer Vision, Natural Language Processing (NLP)
  • Generative AI: Retrival Augmented Generation (RAG), LangChain, Vector Databases, Advanced Chunking & Re-ranking Techniques
  • AI Agents: LangGraph, Google ADK, Google Agent Space, Open AI Agents SDK, Crew AI, Agentic RAGs

Web Technologies
  • HTML, CSS, Java Script, Spring Boot, Microservices, REST APIs, MySQL, Oracle Database, Git, Bitbucket, Agile, Jira, SDLC

Cloud & Deployment
  • Azure DevOps, AWS Services, Azure ML Studio, MLFlow, CI/CD, Docker, Jenkins, Kubernetes

Developer Tools
  • VS Code, Anaconda, Jupyter Notebook, Eclipse, MS Office, Tableau

Projects

InsuCompass - AI Assistant to U.S. Health Insurance

Python, LangGraph, Streamlit, AI Agents, ChromaDB, SentenceTransformers
Developed an Agentic RAG system designed to demystify U.S. health insurance by ingesting data from official sources (CMS.gov, VA.gov) and delivering personalized plan recommendations in conversational style. Orchestrated a multi-agent workflow using LangGraph to deliver context-aware insurance advice, leveraging the Groq API for high-speed LlaMA 3 inference & a local ChromaDB vector store for private user experience. [GitHub] [Demo]
May 2025 - June 2025

ScholarPulse - AI Reserach Assistant

Python, LangChain, Streamlit, AI Agents, ChromaDB, SentenceTransformers
Architected the Advanced RAG system to analyze complex academic papers, providing tailored question-answering, summarization, and code generation through LangChain workflow with advanced prompt engineering for LlaMA-4 Maverick insights and Qwen-2.5 code generation, all within a multi-stage RAG pipeline, reducing research comprehension time by 50%. [GitHub] [Demo]
Feb 2025 - April 2025

Retinal Health Diagnostics - Intelligent CAD System

Python, FastAPI, Streamlit, TensorFlow, Computer Vision, NLP
Conceptualized and built a cutting-edge AI-powered diagnostic tool tailored for identifying retinal diseases like cataracts, macular edema, and diabetic retinopathy across various imaging modalities. Our system achieves an impressive 92% accuracy, automating the diagnosis process and generating detailed reports with clinical recommendations. [GitHub] [Demo]
August 2023 - December 2023

Birthday Greetings App

Java, Spring Boot, JSP, MySQL, HTML, CSS, JS
Developed an innovative web application that transforms birthday wishes into memorable experiences by enabling users to send customized greeting cards straight to their friends' inboxes, ensuring each celebration is both unique and delightful. [GitHub]
July 2018 - September 2018

Research & Development

My research lies at the intersection of machine learning, computer vision, and natural language processing (NLP), with a focus on developing creative, efficient, and responsible AI solutions. I specialize in advancing deep learning methods for multi-modal learning by integrating data from various modalities - such as images, text, and sensor inputs - through scalable attention networks and vision-language models designed for edge deployment. Focused on medical applications, my work aims to develop explainable AI systems that enhance diagnostic accuracy and support decision-making in complex real-world scenarios.

Conference Proceedings
  • 📄{{title}}{{new}}
    {{authors}}
    {{medium}}, {{year}}

Journal Papers
  • 📄{{title}}{{new}}
    {{authors}}
    {{medium}}, {{year}}

Workshop Papers
  • 📄{{title}}{{new}}
    {{authors}}
    {{medium}}, {{year}}

Under Review
  • 📄{{title}}{{new}}
    {{authors}}
    {{medium}}, {{year}}

Research Services

Conference Reviewer

Journal Reviewer

Team

Dr. Dong Hye Ye
Dr. Dong Hye Ye

Advisor during my master's at GSU and TReNDS Center lab.

Dr. Jyostna Devi Bodapati
Dr. Jyostna Devi Bodapati

Advisor during my bachelors's at VFSTR University

Mr. Teja Krishna Cherukuri
Mr. Teja Krishna Cherukuri

Friend, co-author, and colleague from VFSTR, TCS, GSU, and TReNDS Center

Certifications

Interests

Apart from my work and research, I have a diverse range of interests that keep me engaged and inspired. I am passionate about music, with a particular love for classical genre. I find great pleasure in watching television shows and movies, immersing myself in different narratives and perspectives.

I am also an aspiring chef, constantly experimenting with new recipes and culinary techniques. My curiosity and love for learning extend to the digital world, where I spend a significant amount of my free time exploring the latest advancements in Artificial Intelligence and Machine Learning.

Additionally, I enjoy staying active by playing badminton and working out at the gym, which helps me maintain a healthy lifestyle. These activities not only provide a physical outlet but also help me maintain a balanced and well-rounded lifestyle.

Awards 🏆