DVIR COHEN: DATA QUEST
DVIR COHEN
SENIOR DATA SCIENTIST
Accomplished Data Scientist with over 10 years of experience in data analysis, machine learning, and natural language processing (NLP). Demonstrated success in developing and implementing advanced machine learning models across diverse industries, with a focus on enhancing product offerings and user experiences.
EQUIPPED SKILLS
DATA ANALYSIS
DATA ANALYSIS
Expert-level skill in analyzing complex datasets and extracting insights
+50 Data Insight
+30 Decision Power
MACHINE LEARNING
MACHINE LEARNING
Mastery of supervised and unsupervised learning algorithms
+45 Prediction Power
+25 Automation
NATURAL LANGUAGE PROCESSING
NATURAL LANGUAGE PROCESSING
Expert in text analysis, language modeling, and text generation
+45 Text Understanding
+40 Language Generation
GENERATIVE AI
GENERATIVE AI
Work with cutting-edge generative AI models and techniques
+40 Content Creation
+35 Innovation
KEY ACHIEVEMENTS
RESEARCH PAPERS
RESEARCH PAPERS
Published multiple papers in respected computer science journals
+30 Academic Prestige
PATENTS
PATENT
Method and System for Clustering Darknet Traffic Streams with Word Embeddings
+50 Innovation
10+ YRS EXP
INDUSTRY EXPERIENCE
Over 10 years of experience in data science and machine learning
+60 Industry Knowledge
PROJECT LEADERSHIP
PROJECT LEADERSHIP
Worked with cross-functional teams in developing and implementing data science solutions
+40 Leadership
+35 Team Management
PRESS SPACE OR SCROLL TO CONTINUE
COMPLETED QUESTS
PROFESSIONAL JOURNEY
TECH LEAD
DATA SCIENCE TECH LEAD
Wix.com
December 2021 – Present
ACHIEVEMENTS:
Lead the development of data-driven solutions to optimize user engagement and product functionality
Collaborate with cross-functional teams to integrate machine learning models into various platforms
Conduct research on generative AI and contribute to the company's AI initiatives
Develop RAG systems to enhance search functionality and content generation capabilities
EXPERT
CUSTOMER FACING DATA SCIENTIST
Explorium.ai
September 2020 – December 2021
ACHIEVEMENTS:
Designed and implemented end-to-end machine learning solutions for both internal use and external clients
Worked on a variety of use cases, including fraud detection, churn prediction, lead scoring, and lifetime value prediction
Managed the entire product pipeline: exploratory analysis, feature engineering, algorithm development, evaluation, and incorporating customer feedback
SENIOR
SENIOR DATA SCIENTIST
Deutsche Telekom
March 2017 – September 2020
ACHIEVEMENTS:
Identified business problems solvable through analytics and guided team projects from an analytical perspective
Developed prototype models and proofs of concept for various applications
Led the analytic components of products, including exploratory analysis, feature engineering, algorithm development, and evaluation
ROOKIE
DATA SCIENTIST
Deutsche Telekom
January 2017 – March 2017
ACHIEVEMENTS:
Designed a machine learning framework for detecting specific actions in encrypted traffic
Defined and mapped security awareness taxonomy based on related work
Developed a machine learning tool for the automatic creation of fake CVs
SKILL INVENTORY
ABILITIES & EXPERTISE
🧠 CORE COMPETENCIES
GENERATIVE AI
GENERATIVE AI (GPT, Claude, LLaMA, Mistral, Gemini)
Mastery of frontier AI models that create text, code, and beyond.
🪄 +60 Creativity
📈 +40 Productivity Boost
🧠 +30 Cognitive Overload Resistance
LARGE LANGUAGE MODELS
LARGE LANGUAGE MODELS (LLMs)
Command of transformer titans trained on internet-scale knowledge.
🧩 +55 Reasoning
📚 +50 Knowledge Recall
⚙️ +35 Adaptability
PROMPT ENGINEERING
PROMPT ENGINEERING
Shaping your words like spells to guide AI responses.
💬 +60 Command over LLMs
🧠 +40 Few-shot Efficiency
🎯 +35 Response Relevance
RAG
RETRIEVAL-AUGMENTED GENERATION (RAG)
Infusing LLMs with external sources for grounded truth.
🔍 +50 Context Recall
📦 +45 Factual Accuracy
⚡ +30 Latency Optimization
MODEL FINE-TUNING
MODEL FINE-TUNING & INSTRUCTION TUNING
Specializing models with curated datasets and alignment techniques.
🛠 +55 Customization
📐 +50 Task Precision
🔐 +35 Compliance
EMBEDDING MODELS
EMBEDDING MODELS & VECTOR SEARCH
Wielding embeddings to find the signal in the noise.
🧭 +50 Search Relevance
🧱 +45 Retrieval Power
⚡ +25 Speed Boost
MODEL EVALUATION
EVALUATION OF GENERATIVE MODELS
Judging your model like a wise elder—metrics and all.
🎓 +60 Judgment
🛡 +45 Faithfulness
🔍 +40 Insight Extraction
📚 CLASSIC ML
SUPERVISED LEARNING
SUPERVISED LEARNING
Expertise in classification and regression algorithms.
🎯 +50 Prediction
📈 +45 Accuracy
🧠 +35 Generalization
UNSUPERVISED LEARNING
UNSUPERVISED LEARNING (CLUSTERING, DIMENSIONALITY REDUCTION)
Finding hidden structure in unlabeled lands.
🧩 +50 Pattern Recognition
🌀 +40 Exploratory Power
📉 +35 Noise Reduction
TIME SERIES ANALYSIS
TIME SERIES ANALYSIS
Harnessing trends, seasonality, and forecasting.
⏳ +45 Temporal Awareness
📈 +40 Trend Capture
📅 +35 Forecast Accuracy
MODEL EVALUATION
MODEL EVALUATION
Cross-validation, metrics, and test sets—your training dojo.
🧪 +50 Reliability
🔁 +40 Reproducibility
🔍 +30 Metric Mastery
FEATURE ENGINEERING
FEATURE ENGINEERING
Alchemizing raw data into pure predictive power.
⚗️ +60 Signal Boost
🧱 +45 Model Performance
🔧 +30 Customization
🛠 TECHNICAL SKILLS
PYTHON
PYTHON (NumPy, Pandas, Scikit-learn, PyTorch)
Versatile, expressive, and your go-to spellbook for everything.
🐍 +65 Versatility
📊 +50 Data Wrangling
🤖 +45 Model Building
HUGGING FACE
HUGGING FACE
Gateway to a world of ready-made LLMs and embeddings.
📦 +55 Model Access
⚙️ +50 Transfer Learning
💡 +40 Experimentation
LANGCHAIN
LANGCHAIN
Your framework for chaining tools, memory, and logic into LLM agents.
🧱 +60 Modular Design
🧠 +45 Agent Orchestration
📚 +35 Knowledge Integration
AI APIs
OPENAI, ANTHROPIC, COHERE, MISTRAL APIs
Binding with modern AI gods via their sacred endpoints.
⚡ +50 Integration Speed
🌐 +55 Model Flexibility
💰 -10 Cost Mana
STREAMLIT
STREAMLIT FOR APP PROTOTYPING
Turning scripts into sleek, interactive interfaces.
🎨 +50 Visual Appeal
🧪 +45 Rapid Testing
🚀 +35 User Feedback Loops
LANGUAGES
English
ENGLISH
Professional working proficiency
+50 Communication
Hebrew
HEBREW
Native proficiency
+50 Communication
TRAINING & LEARNING
EDUCATION & ACHIEVEMENTS
MASTER
MASTER'S DEGREE IN SOFTWARE AND INFORMATION SYSTEMS ENGINEERING
Ben-Gurion University of the Negev
2018 – 2019
Specialization in Data Science and Machine Learning. Advanced coursework in algorithms, statistical methods, and information systems architecture.
B.A.
BACHELOR'S DEGREE IN COMPUTER SCIENCE
Ben-Gurion University of the Negev
2014 – 2018
Specialization in Big Data. Foundational coursework in computer science theory, software development, and data structures.
KNOWLEDGE SCROLLS
PUBLICATIONS & RESEARCH
METHOD AND SYSTEM FOR CLUSTERING DARKNET TRAFFIC STREAMS WITH WORD EMBEDDINGS
Patent
Developed a system for analyzing and clustering darknet traffic streams using word embeddings, enhancing the detection of malicious activities.
HOW AND WHEN TO STOP THE CO-TRAINING PROCESS
Expert Systems with Applications
September 17, 2021
This research explores optimal stopping criteria for co-training algorithms, providing a novel approach to semi-supervised learning efficiency.
READ SCROLL
ANDROID MALWARE DETECTION VIA AN APP SIMILARITY GRAPH
Computers & Security
June 29, 2021
A groundbreaking approach to detecting Android malware by constructing and analyzing similarity graphs between applications.
READ SCROLL
DANTE: A FRAMEWORK FOR MINING AND MONITORING DARKNET TRAFFIC
European Symposium on Research in Computer Security (ESORICS)
July 7, 2020
Introduced DANTE, an innovative framework for analyzing darknet traffic to detect and monitor malicious network activities.
READ SCROLL
GENERATION OF AUTOMATIC AND REALISTIC ARTIFICIAL PROFILES
International Conference on Advances in Social Networks Analysis and Mining
July 16, 2018
Presented a method for generating realistic artificial profiles that can be used for security research and testing.
READ SCROLL
EMBEDDING-CENTRALITY: GENERIC CENTRALITY COMPUTATION USING NEURAL NETWORKS
International Workshop on Complex Networks
February 15, 2018
Proposed a novel approach for computing node centrality in complex networks using neural network embeddings.
READ SCROLL
CONTACT QUEST
JOIN FORCES WITH DVIR
Interested in discussing data science, machine learning, or potential collaborations? I'm always open to connecting with fellow data enthusiasts, industry professionals, and potential employers. Choose your preferred method of communication to start a conversation!
PRESS ENTER TO CONTINUE
DVIR COHEN: DATA QUEST
© 2023 DVIR COHEN - ALL RIGHTS RESERVED