I'm an AI Engineer & Data Scientist with 2+ years of professional experience building reliable, scalable AI systems that solve real business problems โ from LLM-powered document automation to geospatial ML models.
"Impactful AI must also be trustworthy and responsible."
Currently pursuing a second M.Sc. in Computer Science (Security & Ethical AI) at the University of Passau ๐ฉ๐ช, following a completed M.Sc. in AI Engineering. My thesis explored LLM-driven knowledge extraction and species interaction network reconstruction from scientific literature.
mahshid = {
"focus": ["AI Engineering", "Gen AI", "Agentic AI", "Data Science", "Ethical AI"],
"currently": "M.Sc. Computer Science โ Security & Ethical AI @ Uni Passau",
"day_job": "Data Scientist Working Student @ Next Data Service AG, Berlin",
"languages": ["Python", "R", "SQL"],
"ask_me_about": ["LLM pipelines", "RAG", "FastAPI", "GIS & spatial ML", "AI-Based and Rule-Based Classification"],
}|
Next Data Service AG, Berlin
|
Institut fรผr Integrierte Produktion Hannover
Rah Shahr International Group, Tehran
|
| Degree | Institution | Period |
|---|---|---|
| M.Sc. Computer Science โ Security & Ethical AI | University of Passau, Germany | 2025 โ present |
| M.Sc. AI Engineering โ Applied AI & Software Development | University of Passau, Germany | 2022 โ 2025 |
| B.Sc. Computer Engineering | Ershad Damavand Higher Education Institute, Tehran | 2018 โ 2022 |
M.Sc. Thesis: LLM-Driven Knowledge Extraction and Network Reconstruction of Species Interactions from Scientific Literature
10/2025 โ 01/2026
Automated food hazard and product identification from food safety incident texts using NLP. Tackled real-world class imbalance for food safety monitoring and risk assessment.
10/2025
AI-driven system for automated classification and key-field extraction from multilingual banking documents using LLMs and semantic retrieval. Deployed with MLOps principles via Docker pipelines.
01/2025 โ 07/2025
Designed an NLP pipeline leveraging LLaMA-based models to extract species relationships from scientific texts and construct validated biodiversity knowledge graphs. Achieved F1 = 60%, Precision = 44%, Recall = 91%.
08/2023 โ 10/2023
ML and GIS-based models for optimal turbine site selection using environmental and spatial data. Stack: ArcGIS, QGIS, Python, GeoDjango, Scikit-learn, Geopandas, PostgreSQL.
| Language | Level |
|---|---|
| ๐ฎ๐ท Persian | Native |
| ๐ฌ๐ง English | Fluent |
| ๐ฉ๐ช German | B1โB2, active coursework |
| ๐ฐ๐ท Korean | A2 |
- ๐ค Deepening expertise in LangGraph & agentic AI systems for enterprise workflows
- โ๏ธ Scaling production AI on GCP / Vertex AI at Next Data Service AG
I'm actively looking for full-time roles in AI Engineering, Software Engineering, or Data Science. Open to relocation and on-site work. Let's connect!