Open to opportunities

Ruken Missonnier

@rukenmissonnier

Message

Data scientist specializing in NLP, model fine-tuning, and production-ready ML pipelines.

France

Message

What I'm looking for

I seek NLP/ML roles where I can build and productionize models, work on RAG and LLM fine-tuning, and collaborate in data-driven teams focused on impact.

I am a data scientist focused on natural language processing, model optimization, and deploying robust ML systems to production. I curate and augment multi-source datasets and build retrieval-augmented generation pipelines to power downstream applications.

At Largo.ai I fine-tune and distill large language and transformer models (LLaMA2, RoBERTa, BERT variants, DistilBERT, Qwen-14B) using FP16, LoRA, 4-bit quantization and extensive hyperparameter sweeps to improve accuracy and F1 metrics. I also implemented FAISS HNSW ANN, cross-encoder fusion and async GPT-4 re-ranking for high-quality JSON outputs.

In banking roles I engineered and serialized logistic regression and XGBoost credit-risk models that increased AUC-ROC by 5%, built internal chatbots to query complex SQL data, and deployed ETL pipelines integrated with SAP BusinessObjects. I have experience prototyping anomaly detection models and delivering PowerBI visualizations from PL/SQL and Python pipelines.

I bring academic rigor as a current PhD candidate and prior teaching experience, combining strong statistical foundations with hands-on production experience to deliver impactful, reliable ML solutions.

Experience

Work history, roles, and key accomplishments

Current

NLP Data Scientist

Current

Largo.ai

Jan 2024 - Present (1 year 9 months)

Curated and augmented multi-source emotion/genre and movie/actor datasets and built RAG pipelines with FAISS HNSW, cross-encoder fusion and GPT-4 re-ranking to produce structured JSON outputs. Fine-tuned and distilled LLaMA2, RoBERTa, BERT variants and Qwen-14B using FP16, LoRA and 4-bit quantization with hyperparameter sweeps, improving retrieval and classification metrics (accuracy, F1, top-k).

RAG Faiss GPT 4 Re Ranking LLaMA 2 RoBERTa Python LoRAs

Senior Data Scientist

Emirates NBD & Deniz Bank

Jan 2022 - Jan 2024 (2 years)

Engineered and serialized logistic regression and XGBoost credit-risk models, boosting AUC-ROC by 5% and enabling production deployment; built an internal chatbot to query and retrieve data from complex SQL tables for team use.

Risk Modeling XGBoost Logistic Regression SQL Serialization Chatbot Development AUC ROC Python

IT Data Management Specialist

BNP Paribas & TEB

Jan 2021 - Jan 2022 (1 year)

Deployed ETL pipelines into production integrated with SAP BusinessObjects dashboards and prototyped an ML-based anomaly detection model to flag irregular transaction patterns in accounting and statute-of-limitations data.

ETL Business Objects Anomaly Detection Python Data Integration SQL Data Pipelines

Data Specialist

Türkiye İş Bankası

Jan 2019 - Jan 2021 (2 years)

Built PL/SQL and Python data pipelines to extract and transform raw data and produced Power BI visualizations and reports for team consumption, improving data accessibility and reporting workflows.

Pl SQL Python ETL Power BI Reporting SQL Data Visualization

Statistics Teaching Assistant

Middle East Technical University

Jan 2016 - Jan 2017 (1 year)

Led recitation labs for Statistical Inference and Regression Analysis, supporting over 60 students and assisting with coursework and exam preparation.

Teaching Grading