cv
My academic CV and research experience.
General Information
| Full Name | Mahdi Erfanian |
| Location | Chicago, Illinois, USA |
| Languages | English, Persian (Farsi) |
Education
-
December 2022 - Present Direct Ph.D., Computer Science
University of Illinois Chicago, Chicago, USA - {"Thesis"=>"Generative AI for Multimodal Data Management"}
- M.S. in Computer Science (awarded en route to Ph.D.), Fall 2025
- {"GPA"=>"4.00/4.00"}
-
September 2017 - June 2022 Bachelor of Science, Computer Engineering
Sharif University of Technology, Tehran, Iran - {"GPA"=>"3.70/4.00"}
- Ranked 60th nationwide among ~150,000 participants in the Iranian University Entrance Exam (Konkour)
Research Experience
-
Jan 2026 - Present Ph.D. Research Intern
Microsoft (Code | AI team), Redmond, WA (Remote) - Researching hallucination mitigation in large language models and the use of generative AI as a coding agent (GitHub Copilot, VS Code)
- Designed Delulu, the first execution-verified multilingual benchmark for hallucination detection in fill-in-the-middle code generation
- Investigated synthetic-data fine-tuning for FIM models, improving Qwen-7B exact-match performance by 22% on code hallucination benchmarks
- Proposed an algorithm for multi-turn routing in VS Code and GitHub CLI Auto mode, reducing cancellation rate by 9% and user dissatisfaction rate by 16% in production
- Fine-tuning frontier LLMs on curated datasets to reduce hallucination rates in real-world code completion
-
June 2023 - Present Graduate Research Assistant
InDeXLab, University of Illinois Chicago - Conducting research under Dr. Abolfazl Asudeh on databases, LLMs, and responsible data management
- Developed Needle, a generative-AI-powered text-to-image retrieval framework achieving up to 3x mean-average-precision improvement over OpenAI's CLIP on complex natural-language queries
- Co-developed RSR, an efficient binary/ternary matrix multiplication algorithm, accelerating inference by up to 24x over NumPy and 2.5x on quantized LLMs
- Developed Chameleon, a fairness-aware data augmentation method improving accuracy on under-represented groups by an average of 22% on FERET
- Co-developed FairEM360, a framework for auditing and mitigating bias in entity-matching pipelines
-
December 2022 - August 2023 Graduate Research Assistant
Dreese Lab, The Ohio State University - Researched graph and pattern mining under the supervision of Dr. Srinivasan Parthasarathy
- Developed SYSML+, an enhanced stylometry framework that improved author identification accuracy by 3% over the baseline SYSML system
Industry Experience
-
March 2022 - August 2022 Software Engineer
Software Engineering Lab, Sharif University of Technology - {"B.Sc. Project"=>"Architected and built a CI/CD pipeline for a containerized, microservice-based web application"}
- Optimized a stock market application's deployment from a monolithic to a distributed architecture, improving scalability and deployment speed
- Decreased production Docker image size by over 95% compared to traditional build methods, enabling faster deployments
-
September 2019 - March 2022 Data Engineer
Divar Corp. - Divar is Iran's largest classifieds platform with over 40 million active users and 200 TB of data
- Engineered and maintained the core data pipeline (Airflow, Spark, S3) processing over 200TB of data, improving data availability for a team of 40+ data analysts
- Deployed a distributed JupyterHub on Kubernetes, enabling on-demand, scalable analysis environments for the data science team and cutting down experiment setup time
- Automated ETL deployment workflows using GitLab CI/CD, reducing manual deployment time from hours to minutes and eliminating release errors
-
August 2019 - November 2019 Software Engineering Intern
Rahnema College - Designed and developed an E-commerce auction application using a microservice architecture (Spring Boot, ReactJS, Docker)
- Implemented a push notification service from scratch using Spring Boot and WebSocket
Technical Skills
| Programming | Python, Java, Go, C/C++, Scala, SQL, JavaScript, Bash, Kusto |
| AI / ML | PyTorch, Hugging Face Transformers, NumPy, Pandas, NetworkX |
| Data Systems | Apache Spark, Apache Airflow, HDFS, PostgreSQL, MySQL, Redis, Milvus and other vector databases |
| Cloud & DevOps | Docker, Kubernetes, Azure AI Services, Azure DevOps, GitLab CI/CD, Git |
| Web & Application Frameworks | FastAPI, Django, Spring Boot, React |
Honors and Awards
- VLDB 2026 Student Travel Award (NSF-sponsored), VLDB Endowment
- Top 0.04% nationally in the Iranian University Entrance Exam (Konkour): ranked 60th of ~150,000 participants
- Perfect graduate GPA (4.00/4.00) across Ph.D. coursework at UIC
Open-Source Projects and Contributions
-
Visual Studio Code
Microsoft - Contributor to Microsoft's open-source code editor; landed changes in the GitHub Copilot integration with the Code | AI team
-
Needle and RSR
UIC InDeXLab - Open-source research artifacts for generative-AI image retrieval and binary/ternary matrix multiplication, with code and demos
- RSR achieves up to 24x speedup over NumPy
-
BibTeX Verifier
Independent open-source project (75+ GitHub stars) - Browser-based tool that audits .bib files against Semantic Scholar and Crossref, displays field-level differences, and exports a corrected bibliography
Professional Activities
-
2025 - 2026 Invited Lectures/Talks
University of Illinois Chicago - CS516. Responsible Data Science and Alg. Fairness: Guest Lecture on Generative AI and its applications in Fairness, UIC (2026)
- CS418. Introduction to Data Science: Guest Lecture on Generative AI and Multimodal Data Management, UIC (2025)
-
2018 - Present Teaching Assistant
University of Illinois Chicago, The Ohio State University, Sharif University of Technology - {"UIC"=>"Databases (Spring 24, Fall 23)"}
- {"OSU"=>"Introduction to Java (Summer 23, Spring 23)"}
- {"SUT"=>"Computer Networks (Spring 21), Computer Architecture (Spring 21, Fall 20), Computer Structure and Languages (Fall 20), Systems Analysis and Design (Fall 20), Design of Algorithms (Spring 20), Fundamentals of Programming (Fall 18)"}
-
2024 - Present Program Committee & Reviewing
Various - PC Member: WWW 2026, CIKM 2026/2025/2024, KDD 2026, NeurIPS 2026
- Reviewer: ICLR 2026, NeurIPS 2025 (DynaFront), TKDE 2025/2024, PETRA 2024 (ETHER-AI)
Research Interests
-
Multimodal Data Management
- Text-to-image retrieval systems
- Foundation models for data augmentation
- Vector databases and information retrieval
-
Generative AI and LLMs
- Large language model applications
- Synthetic data generation
- Algorithmic fairness in AI systems
-
Algorithmics and Distributed Systems
- Efficient matrix multiplication algorithms
- Binary and ternary neural networks
- Scalable data processing systems
Notable Courses
- Cloud Computing (A/Grad)
- Advanced Algorithms (A/Grad)
- Introduction to Network Science (A/Grad)
- Algorithms I (A/Grad)
- Software Engineering (4/4)
- Modern Information Retrieval (4/4)
- Functional Programming (4/4)
- Design of Algorithms (4/4)
- Data Structures and Algorithms (4/4)
- Databases Design (4/4)