Data Scientist ML Engineer AI Engineer Full-Stack Machine Learning

На данный момент AI Engineer Data Scientist GenAI.

Мои компетенции и опыт

Data Scientist | ML Engineer | AI Engineer – Full-Stack ML & DL

Data Scientist and ML/AI Engineer with 3+ years of professional experience covering the full spectrum of machine learning and deep learning — classical ML, computer vision, NLP, financial ML, Generative AI, LLMs, and production MLOps.

Strong expertise in: Python, SQL, PyTorch, TensorFlow, LangChain, scikit-learn, XGBoost, OpenCV.

Designed, built, and deployed: production-grade ML models, computer vision systems, LLM pipelines, RAG architectures, agentic AI workflows, and financial ML models that directly drove business impact across regulated and fast-paced environments.

Degree: Master's in Machine Learning. Bachelor's: Statistics and Applied Mathematics. Certification: Deep Learning Specialist — Carnegie Mellon University.

Currently working as AI Engineer – Data Scientist GenAI at Natixis (BPCE Group). Previously worked as Data Scientist / ML Engineer at Adaptive (USA).

Professional Experience:

Throughout my career I have solved a wide range of ML and AI challenges across every major domain of the field, including:

Classical ML & Statistical Modeling

• Regression, classification, clustering, anomaly detection, and time-series forecasting on large structured datasets.
• Statistical analysis, causal inference, hypothesis testing, A/B testing, and experimental design.
• Feature engineering, model validation, and performance optimization (scikit-learn, XGBoost, LightGBM).
• Customer segmentation, behavioral analysis, and churn prediction.

Deep Learning & Computer Vision

• Design and training of deep neural networks from scratch (PyTorch, TensorFlow, Keras).
• Computer vision systems: image classification, object detection, segmentation (CNN, ViT, YOLO).
• Sequence modeling: LSTM, Transformer, time-series deep learning.
• Model optimization, ablation studies, and evaluation under distribution shift.

Natural Language Processing (NLP)

• Text classification, named entity recognition, semantic search, and summarization.
• Transformer-based models (BERT, GPT, LLaMA, Mistral) — fine-tuning and evaluation.
• Embedding pipelines and vector database integration (FAISS, Pinecone, pgvector).

Generative AI & LLMs

• LLM system design, prompt engineering, RAG pipelines, and agentic workflows (LangChain, LangGraph, LlamaIndex).
• Fine-tuning and evaluation of foundation models against real production data.
• Multi-step agent architectures from proof of concept to production deployment.
• Unstructured document processing: PDF, Word, PowerPoint, emails.

Financial ML

• Predictive modeling and risk scoring on financial datasets.
• Credit risk, anomaly detection, and fraud detection in regulated banking environments.
• Time-series forecasting for financial performance and demand modeling.
• Price elasticity, scenario simulation, and statistical attribution models.

MLOps & Production Engineering

• End-to-end ML lifecycle: data processing → training → evaluation → deployment → monitoring.
• CI/CD pipelines, Docker, Kubernetes, model serving, and drift detection.
• Cloud platforms: AWS (SageMaker), Azure ML, GCP (Vertex AI).
• MLflow, DVC, FastAPI, Spark, Hadoop.

Data Analysis & Business Intelligence

• Exploratory data analysis and insight extraction from large, complex datasets.
• Dashboard and report development: Power BI, Tableau, Plotly, Streamlit.
• Translating complex analytical findings into clear business recommendations.
• Cross-functional collaboration with business, engineering, and finance teams.

Experience at Natixis (BPCE Group) — Fins'AIght Program:

As AI Engineer – Data Scientist GenAI, my main responsibilities include:

1. Developing Generative AI prototypes and RAG pipelines on unstructured financial documents (PDF, Word, PowerPoint, emails).
2. Building ML models (scikit-learn, XGBoost) and LLM-based solutions for document intelligence and automated reporting.
3. Statistical analysis of financial data to extract actionable business insights.
4. Automating document and report generation at scale with Python.
5. Contributing to 24 enterprise Data & AI use cases: document intelligence, anomaly detection, predictive modeling, and automated reporting.
6. Working within a large-scale regulated banking environment using Spark, Hadoop, and CI/CD.
7. Collaborating with Data Scientists, Data Engineers, and Finance teams to deliver production-grade AI under real enterprise constraints.

Tech Stack: Python · PyTorch · TensorFlow · Keras · scikit-learn · XGBoost · LightGBM · OpenCV · Hugging Face Transformers · LangChain · LangGraph · LlamaIndex · FAISS · Pinecone · SQL · Spark · Hadoop · Docker · Kubernetes · MLflow · DVC · FastAPI · AWS SageMaker · Azure ML · GCP Vertex AI · R · NLTK · spaCy · Power BI · Tableau · Plotly · Streamlit · Git · CI/CD

Experience at Adaptive (USA) — Data Scientist / ML Engineer:

As Data Scientist / ML Engineer (remote, Dec. 2023 – Mar. 2026), I worked across the full ML and DL stack, delivering production-grade solutions across multiple domains:

1. Designed and deployed end-to-end ML pipelines — classification, regression, clustering, anomaly detection, and time-series forecasting — on large-scale structured and unstructured datasets.
2. Built and shipped LLM systems, RAG pipelines, and agentic AI workflows from proof of concept to live production (LangChain, LangGraph, FAISS, PyTorch, AWS, Azure, Docker, Kubernetes, MLflow). ~40% error reduction, +23% retrieval accuracy.
3. Developed NLP pipelines for text classification, semantic search, named entity recognition, and document processing using Hugging Face Transformers, NLTK, and spaCy.
4. Built computer vision models for image classification and object detection (CNN, ViT, PyTorch, OpenCV).
5. Applied financial ML techniques: predictive modeling, risk scoring, anomaly detection, and time-series forecasting on financial datasets.
6. Implemented causal inference frameworks: difference-in-differences, double ML, A/B testing, and experimental design to measure real-world business impact.
7. Owned full MLOps lifecycle: data processing, feature engineering, model training, evaluation, deployment, monitoring, and continuous improvement (Docker, Kubernetes, CI/CD, MLflow, FastAPI).
8. Communicated findings and model impact through dashboards (Plotly, Streamlit, Power BI) and written reports to technical and non-technical stakeholders.
9. Operated fully autonomously in a fast-paced remote environment across time zones with no oversight required.

Tech Stack: Python · PyTorch · TensorFlow · scikit-learn · XGBoost · LightGBM · LangChain · LangGraph · FAISS · Hugging Face · OpenCV · SQL · Docker · Kubernetes · AWS · Azure · MLflow · FastAPI · Plotly · Streamlit · Power BI