Work Experience & Publications.
This page showcases my work in data science and analytics, covering internships, research projects, and publications that reflect my development across technical and analytical domains. I’ve worked with Python as my primary language, applying machine learning (ML), natural language processing (NLP), and statistical methods to solve real-world problems across business and research contexts. Along the way, I’ve built skills in data wrangling, model development, exploratory analysis, visualization, and stakeholder-focused insight delivery using tools like SQL, Databricks, Scikit-learn, Tableau, and NLP libraries. Each experience helped me strengthen my ability to translate complex data into meaningful outcomes, adapt quickly, and continuously learn. Below is a year-wise timeline linking to the projects and papers that have shaped my journey.
2025
2024
PUBLICATION
Editorial Co-author for Asia-Pacific Journal of Oncology Nursing
Anjali YELLAPUNTULA VENKETA, Sanju RAJAN, Karthik ADAPA
Ongoing as of April 2025
​
ABSTRACT:
This editorial explores how ambient artificial intelligence (AI)—particularly AI scribes—can reduce documentation burdens among oncology nurses in the Asia-Pacific region. Drawing from studies in APAC countries such as India, South Korea, and global implementations, we assess how speech recognition, NLP, and large language models (LLMs) are transforming clinical documentation. The piece highlights both the promise and limitations of ambient AI, including challenges with language diversity, data privacy, and deployment in low-resource settings. We call for inclusive design, rigorous evaluation, and equitable policy frameworks to ensure these tools support—not replace—the human connection in clinical care.​
PUBLICATION
Modeling Workplace Predictors of Depression and PTSD in Healthcare Professionals Using Machine Learning
Anjali YELLAPUNTULA VENKETA, Lukasz MAZUR, Karthik ADAPA
MedInfo 2025 – Full Paper Acceptance (Taipei, Taiwan)
​ABSTRACT:
This study investigates workplace predictors of depression and PTSD among healthcare professionals using traditional machine learning (ML) methods. Survey data were collected from 134 clinicians across four hospitals in the Southeastern United States using validated tools (PHQ-9, PCL-5, and NIOSH WellBQ). We evaluated four classifiers—Logistic Regression, SVM, Random Forest, and Decision Trees—across three feature selection techniques to identify key factors linked to mental health outcomes. ML models achieved high predictive performance (up to 96% accuracy for PTSD), with key predictors including workplace injuries, availability of health programs, life satisfaction, and workplace harassment. SHAP and permutation importance methods provided interpretable insights into feature contributions. These findings underscore the value of explainable ML in identifying actionable workplace risk factors and support the development of targeted interventions to improve mental well-being among healthcare professionals.
INTERNSHIP
The Assessment and User Experience Strategy (AUXS) department at Duke University Libraries focuses on improving library services through data-driven decision-making and user-centered design.
At AUXS, I conducted usability tests, analyzed user behavior, and iteratively refined survey dashboards. My work helped improve staff-facing system interfaces and provided actionable insights that informed key redesign decisions.
INTERNSHIP
Exponentia.ai is a Mumbai-based AI consulting firm that builds enterprise-grade data science solutions across finance, healthcare, and retail.
At Exponentia, I developed scalable ML workflows and leveraged large language models to analyze unstructured business data. My work automated reporting processes, enhanced decision-making speed, and improved the personalization of insights delivered to end users.
2023
In 2023, I completed my undergraduate degree at Veermata Jijabai Technological Institute (VJTI), Mumbai, and began my Master’s in Information Science at University of North Carolina at Chapel Hill. This year was focused on academics and setting the foundation for my work in data science and analytics.
2022
INTERNSHIP
PwC (PricewaterhouseCoopers) is a global leader in consulting and professional services. I worked with the Advisory – Technology Consulting team in Kolkata, a key hub for PwC’s digital and analytics-driven consulting work. There, I built a voice-enabled BI assistant using Python, NLP, and Google Cloud, streamlining audit workflows and cutting report generation time by 50%.
GRADUATE ASSISTANTSHIP
As a Graduate Assistant at UNC Libraries, I worked on On the Books: Algorithms of Resistance project, a digital archive uncovering racially discriminatory Jim Crow laws to support public research and policy reform. The project empowers scholars, journalists, and educators by making systemic bias in historical legal texts both searchable and analyzable.
I used OCR, regex, and NLP tools like SpaCy to extract text and structure metadata from over 100 South Carolina laws, and deployed machine learning models to tag 3,000+ laws from Southern states. This enhanced searchability and revealed patterns in discriminatory legal language.
