👋 About Me

🚀 Machine Learning Scientist
📍 Toronto, Canada
📧 saeed.abbc@gmail.com
🔗 LinkedIn | GitHub

I’m a Machine Learning Scientist specializing in Generative AI, Deep Learning, and NLP, with expertise in Retrieval-Augmented Generation, Agentic Workflows, and LLM Fine-tuning. Combining a strong software engineering foundation with AI expertise, I develop scalable, production-ready AI solutions that drive innovation.

---

🛠️ Skills

• Programming Languages: Python, SQL, R, Java, C++, JavaScript
• ML & AI: Generative AI, Agentic Workflows, RAG, LLM Fine-tuning, Prompt Engineering, Transformers, Deep Learning, NLP, Statistical Modeling
• ML Frameworks: PyTorch, TensorFlow, LangChain, HuggingFace, LlamaIndex, MLflow, Scikit-learn, PySpark
• Software & Tools: FastAPI, Flask, Django, Streamlit, Docker, AWS, Azure, Kubernetes, Git, Bash Scripting
• CS Fundamentals: Data Structures & Algorithms, Distributed/Parallel Computing, Object Oriented Design Patterns

---

💼 Experience

Machine Learning Researcher

📍 York University Data Mining Lab (May 2022 – Dec 2024)

• Developed a sliding-window document processing method, improving Transformer-based text segmentation F1 scores by 10%.
• Enhanced neural question generation by integrating domain-specific knowledge graphs, improving relevance by 3%.

AI Software Developer

📍 iNAGO Inc. (May 2023 – Aug 2023)

• Optimized RAG pipelines for long-form QA, improving performance by 5% over prompt-tuned LLM baselines.
• Built and deployed scalable REST APIs using FastAPI and Azure, enabling high-throughput inference for NLP models.

Back-end Software Developer

📍 University of Tehran Science & Technology Park (Oct 2019 – Mar 2021)

• Developed real-time graph analysis services using Flask/Django, optimizing performance with native SQL/Neo4j queries.
• Optimized large-scale graph algorithms (centrality, community detection) to enhance big data processing efficiency.

---

🎓 Education

MSc in Computer Science

📍 York University (2021 – 2024)

BSc in Computer Engineering

📍 University of Tehran (2016 – 2021)

---

📜 Publications

WeSWin

• Saeed Abbasi, Aijun An, Heidar Davoudi, Ron Di Carlantonio, and Gary Farmaner. 2025. Neural Document Segmentation Using Weighted Sliding Windows with Transformer Encoders. In Proceedings of the 31st International Conference on Computational Linguistics: Industry Track, pages 807–816, Abu Dhabi, UAE. Association for Computational Linguistics.

---

🔥 Projects

MapEcho

• AI-powered geospatial assistant for interactive querying and real-time visualization of transportation data.
• Designed multi-agent workflows for querying open-source and structured datasets.

LLM-Based Text Segmentation

• Built an interactive web app for real-time text segmentation using LLM embeddings & Gradio.

Insurance Claim Prediction

• Developed an ML pipeline for vehicle insurance claim prediction using Neural Networks & XGBoost.

---

📫 Get in Touch

I’m always excited to collaborate on cutting-edge AI projects. Feel free to reach out via LinkedIn or GitHub!