Naveed Afzal

Research and Development related to Knowledge Graph Construction and Large Language Models.

• Utilize LLMs with Graph RAG for generating news recommendation explanations
• Built a data pipeline for Knowledge graph construction from unstructured texts
• Develop demo applications utilizing Knowledge Graphs for end product use cases
• Developed models and datasets for Entity Relation Extraction based on Distant Supervision
• Contributed to our in-house Product for PDF and Image Document Structure Extraction
• Organized research collaboration activities with a partner university

Responsible for developing new features and improving functionality for existing NLP products.

• Built a prototype of a Question Answer extraction system from audio transcripts
• Created Proof-of-Concept of a call classification system for client companies
• Led annotation of a Dialog Act Corpus and developed a Dialog Act classification model
• Led annotation of a Named Entity Recognition Corpus and developed an NER model
• Created a tool for sentence segmentation, alignment, and punctuation of transcribed audio
• Developed a sentence ranking model applied to text summarization

Did preliminary research and prototyping for various projects, working directly under the CTO.

• Designed a chatbot to be integrated in existing product
• Developed a proof-of-concept for a product search and recommendation app
• Created a system for Time Series Prediction and Trend analysis of walk speed data

Working as an independent contractor, I helped improve the accuracy of a semantic search system for software requirement specifications.

• Built a mechanism for automatic parameter tuning
• Defined a ranking-based scoring mechanism for evaluating system accuracy
• Facilitated preparation of training data and utilized it for parameter tuning and evaluation
• Conducted basic experiments of transfer learning

Advanced studies in computer science with focus on machine learning and natural language processing.

Comprehensive engineering education with strong foundation in mathematics, physics, and electrical systems.

Built a data pipeline for Knowledge graph construction from unstructured texts using LLMs and Graph RAG

Developed models and datasets for Entity Relation Extraction based on Distant Supervision

Built a prototype system for extracting question-answer pairs from audio transcripts

Improved accuracy of semantic search system for software requirement specifications with automatic parameter tuning

"Modularity Optimization as a Training Criterion for Graph Neural Networks"

Complex Networks IX: Proceedings of the 9th Conference on Complex Networks, CompleNet 2018

"Generating Explainable Recommendations of Documents through Personalized Knowledge Graphs Based on User Activity"

YANS 2024