Experience & education
My path spans industry AI engineering, independently built AI systems, graph ML research, and mathematics. Research and postdoctoral appointments overlap in time, as recorded in my CV.
Professional work
Sole AI engineer for a production-facing musculoskeletal computer vision workflow. Designed multi-view detection, pose estimation, segmentation, landmark refinement, geometric measurement, and quality-control components; collaborated with medical experts and supported API integration using a FastAPI/Swagger workflow.
Industrial case study →Implemented a modular RAG system with document ingestion, vector indexing, semantic retrieval, configurable inference, retrieval evaluation (Hit Rate@K, Recall@K, MRR), tests, logging, latency tracking, and reproducible configuration.
Open-source engineering project →Developed and benchmarked ML and deep learning models for graph representation learning and molecular property prediction. Co-developed FunQG, explored knowledge-graph embeddings and link prediction with PyKEEN, and supervised and advised multiple M.Sc. students.
Graph neural networks and deep learning for healthcare-oriented representation learning, predictive modeling, and explainable AI. This period overlaps the MLGM Lab appointment.
Computational algebraic graph theory, graph invariants, combinatorics, and mathematical modeling.
Education
Postdoctoral Research · Machine Learning & Graph AI
Shahid Beheshti University · May 2021 – April 2022.
Ph.D. in Mathematics · University of Isfahan
September 2014 – December 2019 · Graduated with Distinction
GPA: 19.25/20 (approximately 3.85/4.0) · Thesis grade: 20/20.
Teaching & communication
Invited Lecturer (graduate-level Graph Neural Networks & Machine Learning), Teaching Assistant (Applied Machine Learning), and Co-Instructor (graduate-level Optimization for Data Science) at Shahid Beheshti University. Supervised and advised multiple M.Sc. students in Knowledge Graphs, GNNs, and LLM-based AI.
Languages: Persian (native); English (professional working proficiency, CEFR B2+).
Teaching details →