Open to opportunities

Javidan User

@javidanuser

Message

Senior Data Scientist building production ML systems from dirty data to deployed, automated solutions.

Azerbaijan

Message

What I'm looking for

I’m looking to build production-ready ML and NLP systems end-to-end—working with messy data, deploying APIs, and automating retraining to deliver measurable operational savings and faster decision workflows.

I’m a Data Scientist who’s drawn to problems without obvious solutions, especially where the data is “ugly and dirty.” Over 3+ years across oil & gas, banking, and government, I build production ML systems end-to-end—from raw data and feature engineering to deployed APIs and automated retraining.

In my roles at SOCAR|Caspian Innovation Center, I helped deliver tangible impact: production ML systems saved $280K/year in maintenance costs, accelerated document review by 6×, and surfaced 140K previously unknown customers. I’ve also worked on Energy AI using linear optimization to balance plantwide steam and energy networks (users can build plant P&ID in 2 hours), and Should Cost material price prediction that reduced manual review from 2–4 weeks to 15 minutes.

I also focus on scalable delivery and clean engineering practices—standardizing shared DS tooling, improving experiment tracking and versioning, and deploying robust pipelines (e.g., NER model pipelines and MLOps flows). On the research side, I’ve worked with Knowledge Distillation on Whisper and contributed Azerbaijani NLP data for open-source use.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

SOCAR Caspian Innovation Center

Apr 2025 - Present (1 year 4 months)

Built and deployed energy-optimization and prediction systems, including linear optimization-based balancing for plant energy networks and a should-cost material price prediction product. Developed supporting DS tooling (standardized feature engineering modules, experiment tracking, dependency management) and deployed multiple models with automated backends.

Optimization fastAPI Git Docker Trino Feature Engineering Economic Indicators MLFlow ClearML UV развертка Ruff MkDocs Pytest Airflow

Data Scientist

EiGroup Digital Research Laboratory

Jul 2024 - Apr 2025 (9 months)

Designed and deployed a custom scalable NER model pipeline for a governmental body, accelerating the product review process. Worked on oil well parameter prediction using pressure transient analysis and mentored deployment practices.

NLP Minio RabbitMQ SQL Second Derivative Methods

Junior Data Scientist

Kapital Bank

Oct 2023 - Jul 2024 (9 months)

Developed an end-to-end Azerbaijani text-to-speech model and supported its deployment for advertisements and a call center. Built predictive maintenance for ATMs, identified unknown business owners (micro lead prediction), and implemented dynamic data lineage visualization.

Text to Speech (TTS)Whisper Predictive Maintenance Machine Learning Micro Lead Prediction Data Lineage D3.Js HTML

AI Researcher

AI Research Center at Baku Higher Oil School

Aug 2023 - May 2024 (9 months)

Applied knowledge distillation (teacher-student) to a Whisper model using HuggingFace, producing a smaller target model variant. Collected Azerbaijani NLP data for open-source use.

Knowledge Distillation Whisper NLP Data Collection Teacher Student Training