aidyn kadyr

AI & Data Engineer

Yuno.ai S.r.l.

Built a governed BigQuery platform and ingestion pipelines spanning UK and EU company records, with 6M+ company and 60M+ financial and shareholder records.

• Architected and own a BigQuery + Dataform data platform implementing Medallion architecture (Bronze/Silver/Gold layers), unifying the UK company register, iXBRL financial filings, and beneficial-ownership (PSC) data into a governed layer of 6M+ company records and 60M+ financial and shareholder records, with full data lineage and versioned .sqlx transformation definitions (Dataform — BigQuery-native dbt-equivalent).

Data Engineer (Intern)

Infineon Technologies AG

A semiconductor manufacturer.

Automated data workflows and reporting, cut quote retrieval time 10x, and built demand forecasting models.

• Implemented end-to-end CI/CD and data pipelines using Python, SQL, and Tableau, accelerating data transformation workflows and automated reporting by 50% through BitBucket and Jenkins integration. • Generated a Python algorithm to consolidate and analyze historical quote data for 1,200+ customers and 7,000+ files, improving quote retrieval time by 10x and facilitating decision-making for product management and marketing teams. • Built time-series demand forecasting models on historical sales data, enabling supply chain and product teams to optimize inventory planning and reduce excess stock.

Data Scientist (Intern)

UniCredit S.p.A

A European banking group.

Built an internal revenue forecasting model and raised test coverage from 30% to 85% while streamlining financial data synchronization.

• Developed an internal revenue forecasting model in Palantir Foundry using Python and PySpark (Apache Spark), following Agile methodology to improve cost management and financial decision-making. • Implemented TDD frameworks and automated pipelines increasing testing coverage from 30% to 85%, while streamlining financial dataset synchronization from DWH into Palantir.

Machine Learning Engineer

Zypl.ai Inc.

Developed and backtested trading models for illiquid forex pairs, reporting 32% simulated ARR over a 10-year historical backtest.

• Developed classification and regression models for trading illiquid forex pairs, running time-series analysis and backtesting to inform feature selection, achieving 32% simulated ARR across a 10-year historical backtest.

Master of Science, Data Science

Bocconi University

Bachelor of Science, Mathematics

Nazarbayev University