Pj Sheini Email & Phone Number

Vienna, Va, Us

• Architected and developed a question-answering application leveraging Generative AI, including Large Language Models (LLM) through OpenAI API and Retriever-Augmented Generation (RAG), enhanced with citation generation for AI Search Platform solutions• Directed a cross-functional team in a Document Processing and Prompt Engineering project, extracting and manipulating data from PDFs for improved search result relevancy, utilizing ML frameworks like PyTorch and tools like Langchain and LlamaIndex• Cultivated AI/ML development lifecycle skills, including fine-tuning transformer models and Generative AI applications, backed by proficient use of AWS and MLFlow for robust statistical modeling and AI pipeline development• Spearheaded the design and deployment of an AI-powered chat application and semantic search-driven patient chart analysis tool for healthcare providers, focusing on optimizing AI/ML capabilities for scalability and quality

Contractor for Conversational AI Banking team @ Navy Federal Credit Union• Conducted in-depth prioritization of Key Performance Indicators (KPIs) by analyzing their correlations with customer sentiments to drive strategic decision-making.• Designed and built a categorization model for conversations using a Large Language Model (LLM), enabling real-time monitoring and identification of key customer concerns.• Implemented a granular analysis framework for the Microsoft Conversational Language Understanding (CLU) model, facilitating the tracking and enhancement of virtual assistant capabilities.• Executed numerous ad-hoc analyses based on specific requests, delivering actionable insights and data-driven solutions.• Fostered team growth and knowledge sharing by mentoring colleagues in the field of Natural Language Processing (NLP), leveraging domain expertise to empower and upskill team members.

Washington, D.C., N.W., Us

• Taught a graduate-level course titled "Business Data Networks and Security," covering key topics in datacommunication, network design, cybersecurity principles, and business applications of data networks.• Delivered comprehensive course materials, including lectures, assignments, and assessments, to enhancestudents' understanding and practical skills in data networks and security.• Mentored and guided graduate students on academic and professional development, fostering critical thinkingand problem-solving abilities in the context of business data networks and security.

Newbury, Berkshire, Gb

• Submitted a patent (granted Aug 2024) for automating the preprocessing of complex logs, enhancing efficiency in data handling.• Developed a comprehensive data preparation process, including log parsing, ETL, and feature extraction and selection techniques, streamlining data readiness for advanced analytics.• Successfully implemented and deployed machine learning and deep learning solutions for log analysis and anomaly detection tools, significantly improving decision-making for ArcSight's support team. These tools resulted in a substantial reduction in investigation time and enhanced operational efficiency by an order of magnitude.• Prepared presentation materials and formal business case documents for senior management, effectively promoting findings and driving science-based decision support recommendations derived from the successful implementation of machine learning and deep learning solutions for log analysis and anomaly detection tools.• Characterized hostnames and built a model to compute a risk score for each hostname, achieving a 30% reduction in false positives and improving overall accuracy by more than 27% compared to the legacy approach when predicting malicious hostnames.•

Boulder, Colorado, Us

• Led research and development efforts for pathogen surveillance software on persistent memory (Optane) – funded by Intel Corp in collaboration with Columbia University• Characterized traditional techniques for pathogen surveillance and analyzed trends for optimized strategies• Developed a novel DNA representation using self-supervised Learning (SSL) and information retrieval trajectories• Utilized PySpark and SQL to process and analyze large-scale genomic data sets for pathogen surveillance, improving efficiency and scalability of data processing pipelines.• Developed a novel k-mismatch similarity search by implementing a hamming weight tree data structure.•

Tuscaloosa, Al, Us

• Designed and implement a novel method for generating cross-lingual word embeddings using graph theory and shallow neural networks to reconstruct linguistic contexts of words•. Investigated the aspects of individuals revealing personally identifiable information (PII) on social media (Twitter) during a disaster event•. Developed sentiment analysis, information retrieval, semantic search, and text classification applications to benchmark the performance of multilingual word embeddings.• Led and participated in multiple cross-functional projects utilizing custom language models for topic modeling, sentiment analysis, and automatic labeling of Twitter data.•.

Bowling Green, Ky, Us

•Deployed DDoS attack strategies and set up various hypervisors in controlled environments to collect data, enabling robust testing and analysis.• Analyzed data to benchmark and compare each hypervisor's performance under different DDoS attack scenarios, identifying optimal configurations for enhanced security and resilience.• Implemented research objectives, provided research support, and mentored undergraduates•.

Louisville, Ky, Us

• Designed, implemented, and consulted the company employees about “Disaster Recovery Procedures and Protocols” for The Data Center• Performed full spectrum CNO. Assessed cybersecurity vulnerabilities (CNA & CNE) and implemented a mitigation solution (CND) tailored for the business. • Provided technology planning, design, and administration for The Data Center and its network infrastructure• Supported more than 40 remote web servers• Identified the requirements and implemented the most efficient and cost-effective solutions for the resilient network system.

Doctor Of Philosophy (Ph.D.), Computer Science - Nlp

Master'S Degree, Computer Science - Cloud Security