Lefteris Andritsos
Verified Expert in Engineering
Data Scientist and Developer
Athens, Central Athens, Greece
Toptal member since October 21, 2024
Lefteris is an accomplished senior data scientist with over five years of experience developing innovative machine learning solutions. At VMO2, he led the creation of a flagship product recommendation engine, while as a Toptal core member, he focused on productionizing AI models. With a strong background in product and research roles, Lefteris excels at transforming complex data into actionable insights and pushing the boundaries of AI and data science.
Portfolio
Experience
- Research - 13 years
- Machine Learning - 12 years
- Python 3 - 10 years
- SQL - 4 years
- Google Cloud Platform (GCP) - 3 years
- AI Chatbots - 2 years
- Retrieval-augmented Generation (RAG) - 2 years
- ChatGPT - 2 years
Availability
Preferred Environment
Linux, Google Cloud Platform (GCP), Azure, Jupyter Notebook, Vim Text Editor, Python, SQL, ChatGPT, Deepgram, Docker
The most amazing...
...thing I've done is utilize state-of-the-art algorithms to create simple products.
Work Experience
Senior Data Scientist
Toptal
- Designed and implemented an AI-driven audio interview system, surpassing the candidate assessment quality of competitors and enhancing the quality of hiring decisions for client projects.
- Developed a price recommendation engine for talents, leading to increased retention rates and client satisfaction.
- Conducted in-depth research on AI infrastructure, focusing on retrieval-augmented generation (RAG) capabilities. This led to establishing best practices that improved system performance and scalability, benefiting multiple client deployments.
- Developed an AI chatbot assistant equipped with RAG capabilities and coordinated swarms of AI clients, enhancing user interactions and providing personalized responses across multiple applications.
Data Scientist
Virgin Media
- Developed a flagship customer personalization engine utilizing advanced machine learning algorithms, enhancing user experience and driving increased upsell and cross-sell opportunities.
- Implemented a fraud detection system leveraging behavioral forecasting models, successfully capturing fraudulent activity with large savings in potential losses.
- Headed marketing campaign analyses using purchase propensity and customer intent models, optimizing targeting strategies and contributing to an increase in Facebook campaign effectiveness and conversion rates.
- Provided direct support to MLOps by developing pipeline templates on Vertex AI, streamlining deployment processes, and enhancing the efficiency of machine learning workflows.
Data Scientist
Faculty
- Deployed a high-performance CatBoost model on the Google Cloud Platform (GCP), heavily reduced processing time, and increased the model's recall by 50%.
- Conducted A/B testing on a high-traffic product page, showing increased revenue with the new model.
- Completed an intensive 2-week training on advanced machine learning and statistical modeling, strengthening expertise in business needs.
Postdoctoral Researcher
University of Surrey
- Researched and analyzed novel battery materials for electric vehicles, aiming to improve energy capacity and extend battery life through advanced data-driven modeling techniques.
- Published my first machine learning-related publication.
- Collaborated with cross-functional teams from academia and industry, driving innovative solutions for battery technology advancements with machine learning research.
Postdoctoral Researcher
King's College London
- Conducted innovative research on lightweight materials for the automotive industry, enhancing product quality and manufacturing efficiency.
- Secured over £170,000 in grant funding from UK bodies through successful proposal writing, supporting advanced research initiatives in material science.
- Served as Linux system administrator for a local computing cluster, managing system performance and ensuring optimal operation for high-performance computing tasks.
- Applied cutting-edge mathematical and statistical models to simulate material properties, enabling accurate predictions for improving material performance in automotive and aerospace industries.
- Provided research guidance and mentorship to PhD students, enhancing their projects through expert advice on methodologies and data analysis.
Experience
Product Recommendation Engine
ACHIEVEMENTS
• Leveraged CatBoost for gradient boosting and recurrent neural networks (RNN) to perform multi-label classification, effectively predicting relevant product combinations tailored to individual customer profiles.
• Deployed the solution on GCP, utilizing scalable infrastructure to handle large volumes of data and perform batch processing of customer interactions.
• Designed the model to classify multiple products per customer, enabling personalized recommendations based on behavior, preferences, and purchase history. This approach maximized customer engagement and increased the relevance of product offerings.
OUTCOMES
• Achieved a significant increase in upsell and cross-sell rates, contributing to a revenue uplift within the first three months post-launch.
• Enhanced customer satisfaction and engagement metrics, resulting in improved user retention and loyalty by offering personalized model explainability.
Education
Diploma in Manufacturing and Product Design
City Business School - London, UK
PhD in Computational Physics
Queen Mary, University of London - London, UK
Bachelor of Science in Materials Science
Univeristy of Patras - Patra, Greece
Skills
Libraries/APIs
CatBoost, Scikit-learn, NumPy, Pandas, OpenAI API, REST APIs, TensorFlow, PyTorch, LSTM, SciPy
Tools
Jira, Net Promoter Score (NPS), ChatGPT, Git, GitLab CI/CD, GitHub, Apache Airflow, dbt Cloud, MATLAB, Vim Text Editor, Composer, Claude, ARIMA
Languages
Python 3, SQL, Python, Fortran
Paradigms
Agile, High-performance Computing (HPC), Anomaly Detection
Platforms
Google Cloud Platform (GCP), Jupyter Notebook, Vertex AI, Linux, Kubernetes, Docker, Azure
Frameworks
Streamlit, LlamaIndex
Storage
SUSE
Other
Programming, Research, Mathematics, Physics, Machine Learning, Simulations, Statistics, Quantum Physics, Materials Science, Innovation, Grant Proposals, Stochastic Differential Equations, Chemistry, Batteries, ML Pipelines, Natural Language Processing (NLP), Large Language Models (LLMs), OpenAI GPT-4 API, Deepgram, Dify, AI Chatbots, Retrieval-augmented Generation (RAG), Churn Analysis, Neural Networks, A/B Testing, Deep Neural Networks (DNNs), Poetry, Recurrent Neural Networks (RNNs), Data Science, Artificial Intelligence (AI), APIs, Prompt Engineering, ChatGPT API, ChatGPT Prompts, Deep Learning, AI Agents, Chatbots, Manufacturing, Product Research, Big Data, Multimodal GenAI, Stable Diffusion, Back-end Admin Systems, Time Series, API Integration, Directed Acrylic Graphs (DAG), Hugging Face, LangChain, Conversational AI, Speech to Text, Text to Speech (TTS), Battery Management Systems, Dagster, Looker Studio
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring