Looking for a job

Andrei Ivanov

@andreiivanov3

Message

I turn lending and customer data into clearer product decisions using SQL, Python, and experimentation.

Kyrgyzstan

Message

What I'm looking for

I'm open to international and relocation opportunities where I can turn customer, lending, and product data into clear decisions using SQL, Python, experimentation, and machine learning.

I'm analyzing customer flows, lending journeys, and business-rule logic at Koronatech Belarus to identify friction points and product opportunities. I use SQL and Python to validate customer data and translate business needs into analytical requirements.

Previously at Sberbank, I supported commercialization readiness for six machine-learning models across lending and customer communications. I built preprocessing, anonymization, and data-hashing pipelines while monitoring model quality and aligning metrics, risk controls, and acceptance criteria with stakeholders.

I've also fine-tuned and deployed Whisper Large, reducing GPU utilization by 30% while reaching WER ≤ 8%, and built audio-prediction models using CNNs and classical ML. I enjoy making complex data understandable and useful for digital products.

Experience

Work history, roles, and key accomplishments

Current

Junior Data Analyst

Current

Koronatech Belarus

Mar 2026 - Present (6 months)

• Analyze customer flows and business-rule logic across loan issuance and product upgrades.
• Identify friction points, behavioral patterns, and product opportunities in lending journeys.
• Translate business needs into analytical requirements for new products and service changes.
• Use SQL and Python to extract, analyze, and validate lending and customer data.
• Coordinate with business, product,

Data Scientist

Sberbank

Aug 2025 - Feb 2026 (6 months)

• Supported commercialization readiness for six models: five classical machine-learning models and one LLM-based model.
• Validated and monitored model performance; aligned business requirements, metrics, acceptance criteria, and risk controls with stakeholders.
• Analyzed model-driven lending and customer-communication use cases, including hard collection, recommendations, look-alike, response, a

Junior Machine Learning Engineer

Novosibirsk Institute of Program Systems

Apr 2025 - Aug 2025 (4 months)

• Prepared, labeled, and augmented datasets for ASR model training.
• Fine-tuned and deployed Whisper Large; reduced GPU utilization by 30% and achieved WER ≤ 8% on the target dataset.
• Built ML pipelines and developed Transformer- and LLM-based research prototypes.

Research Machine Learning Engineer

Advanced Engineering School of Novosibirsk State University

Nov 2024 - Feb 2025 (3 months)

• Developed CNN models for audio-based outcome prediction.
• Benchmarked XGBoost, LightGBM, and CatBoost against neural-network approaches.
• Evaluated CNN and classical-ML cascades and ensembles; performed hyperparameter tuning, feature selection, and robustness analysis.