Taha Elahibakhsh
AI & LLM Researcher · Agentic Systems
Skills - Education - Languages - contact
About
Passionate AI researcher specializing in data analyse and natural language processing and agentic systems.
Experienced in developing intelligent solutions for petrochemical and financial sectors.
AI Olympiad
Petrochemical AI
Agentic Finance
Adobe Suite
Technical Skills
Python / AI Libraries
PyTorch, TensorFlow, Hugging Face, Transformers, scikit-learn
NLP & LLM models
Language models, embeddings, sequence modeling, tokenization pipelines
Agentic Systems & RL
Multi-agent frameworks, reinforcement learning, autonomous decision systems, N8N, Langchain
Data Science & Analyse
Pandas, NumPy, statistical modeling, data visualization, feature engineering
Front-End Development
HTML, CSS, JavaScript, React, Vue.js, Bootstrap, SASS
Back-End Development
Node.js, Express, Django, Flask, SQL, NoSQL (MongoDB)
Programming Languages & Tools
Python
R
JavaScript
SQL
Git
Docker
C#
Experience & Collaborations
2022
NODET Allameh Helli 7 High School
Admitted to the National Organization for Development of Exceptional Talents (NODET) — Software Engineering major
2022
IOAI Artificial-Intelligence
Artificial-Intelligence Olympiad
2024
ELECOMP Exhibition
Presented our AI-driven crypto prediction and consulting system, gaining investor interest and partnerships with exchanges.
2023
Financial AI Systems Development
Built agentic systems for automated trading analysis
2024
AI Research Collaboration - Petrochemical Industry
Developed predictive models for process optimization
2024
Finalist in Data Science Olympics - Quera
Ranked 21st nationally after completing two data science tasks in the final stage.
2025
7th Place in AI Security Hackathon – Attack & Defense on LLMs
Ranked 7st nationally after completing 3 stage of Secure AI Hackathon and a problem-driven national competition focused on Large Language Model Security.
Selected Projects
NLP
Natural Language Processing System
Advanced NLP pipeline with Petrochemical collaborative for document analysis
RL
Reinforcement Learning & Agentical Trading Agent
Autonomous trading system using deep RL for financial market analysis
ML
Machine Learning Prediction Models
Prediction Models on Dataset with more than 80% accuracy
ML
Deep Learning Prediction Models
DL Prediction Models on Dataset with more than 80% accuracy
ML
AI & LLM secutity model and tasks
Marcus Guard Agent is designed to defend LLM systems against malicious or adversarial prompts.
