About me

Master’s student in Artificial Intelligence with 2+ years of hands-on experience in NLP, deep learning, LLMs, RAG systems, and knowledge graphs. Strong foundation in algorithms, data structures, software engineering, and problem solving, complemented by research experience in LLMs and knowledge sources. Passionate about learning and experimenting with emerging AI technologies to contribute to impactful software and AI projects in industry.

Selected Projects

Conversational Search System using RAG

Developed a retrieval-augmented conversational agent using transformer embeddings on the CMU Book Summaries dataset, integrating a vector DB with hybrid semantic retrieval for relevance and performance evaluation.

Real-Time Collective Sentiment Analysis System for Twitter Streams

Built a real-time system to analyze and visualize collective sentiment across multiple Twitter streams for multiple end users using Apache Kafka and TinyBERT.

Taxonomy Discovery in Ontologies Using Large Language Models

Designed a classification-based system for taxonomy discovery with fine-tuned BERT-Large, demonstrating that our system achieved performance comparable to the much larger LLaMA 3 70B.

• 1st place, LLMs4OL 2024 Challenge Task B (Geonames Ontology)
• Paper published in Open Conference Proceedings

Persian Paraphrase Detection using Deep Learning

Fine-tuned ParsBERT on a subset of the Degarbayan-SC dataset for paraphrase detection, achieving an F1-score of 98.4% on Degrabayan-SC and 90.1% on ExaPPC, showcasing the model’s effectiveness and cross-dataset generalization.

Publications

• Hashemi, S. M. H. et al. “SKH-NLP at LLMs4OL 2024 Task B: Taxonomy Discovery in Ontologies Using BERT and LLaMA 3.” Open Conference Proceedings, 2024. (link)
• Shamsfard, Mehrnoush et al. “FarsEval-PKBETS: A new diverse benchmark for evaluating Persian large language models.” arxiv, 2025. (link)
  
  

Technical Skills & Languages

Programming Languages

• Python
• Java
• SQL (PostgreSQL, SQLite)
• Others: C, Scheme, Verilog

Machine / Deep Learning

• NumPy
• Pandas
• Matplotlib
• Scikit-learn
• PyTorch
• Hugging Face Transformers

NLP / LLM / RAG Tools

• NLTK
• Hazm
• LangChain
• OpenAI API
• LanceDB

Backend Development

• Django
• Spring Boot
• Hibernate
• Flask

DevOps

• Docker
• Linux
• Git

Others

• ElasticSearch
• Apache Kafka
• RDFLib

Languages

• Persian (Native)
• English (Advanced)
  
  

Contact

• Email
• Linkedin