Portfolio

Halina Lukashenko
SQL developer and AI product builder

20+ years of SQL, databases, regulatory reporting and data quality, from a Ukrainian bank to the Danish agricultural sector. Today I also build and test AI tools: Telegram bots, website chat assistants and dashboards, using Gemini and Claude Code.

  • SQL
  • Oracle
  • PL/SQL
  • T-SQL
  • MS SQL Server
  • Data quality & QA
  • Gemini API
  • Claude Code
  • Telegram bots

HL22111770@GMAIL.COM  ·  Tlf. +45 91 92 78 88  ·  LinkedIn

Halina Lukashenko

SQL & data experience

Work for employers; code is not public, details in the CV and references

20%

Data quality fix in a live register

Joining two history tables on ID alone multiplied rows in historical data — a cross-join between periods. About 20% of the rows (1,238 of 5,894) were false. I added a period-overlap condition:

AND o.start_dt <=
  NVL(h.end_dt, DATE '9999-01-01')
AND h.start_dt <=
  NVL(o.end_dt, DATE '9999-01-01')

After the fix, views and SELECTs returned clean data. The condition is now an internal standard: it is required in every join with this table, and the team has adopted it in its procedures, views and functions. For functions it was critical, and the only reliable fix: a function must return exactly one value, and duplicate rows from the join made it fail with a “too many rows” error. I also replaced ROWNUM < 2, which picked an arbitrary row, with ORDER BY … FETCH FIRST 1 ROW ONLY, so the query always returns the right row.

Landbrug & Fødevarer · Oracle

2 min → 1 s

Query performance

Diagnosed and solved performance problems across SQL, PHP and the front end; one query went from 2 minutes to 1 second.

Landbrug & Fødevarer · Oracle

~100

Database objects for research

Developed and adapted about 100 Oracle database objects. Central developer in the OVERLEVELSE (Survival) research project: two complete model variants; the chief scientist called the work "brilliant" and "excellent tables" and thanked me for my "extensive work"; the tables and reports were used in two research presentations.

Landbrug & Fødevarer · Oracle, PL/SQL

22 years

Regulatory reporting in banking

Mandatory reporting to the National Bank of Ukraine and the Deposit Guarantee Fund. Built the first tables and procedures that all later reporting grew from; Deputy Head of the Reporting Directorate, later Head of Area in the Analytical Reporting and Data Warehouse Directorate.

Prominvestbank, Kyiv · Oracle, MS SQL, Sybase (Transact-SQL), Borland C

T-SQL, Python (basic)

Data loading, log reports and a Python migration

T-SQL for loading CSV files into a database and for reports based on log files. Separately, I migrated PowerShell scripts to Python for Eloomi, the municipality's course portal with about 30,000 users.

Aarhus Municipality · MS SQL, T-SQL, Python

AI products

Built and running, with real screenshots

Prototypes & concepts

Ideas and clickable demos
NordPuls fitness chatbot demo page
Clickable demo · in Danish

NordPuls: chatbot for fitness centres

Interactive demo of a website chat assistant: answers opening hours and prices for the right centre, books classes and sells shakes, 24/7. The demo runs on prepared answers; the full version would use an LLM.

Tech: HTML · JavaScript

Sundt Liv health food store chatbot demo page
Clickable demo · in Danish

Sundt Liv: chatbot for health food stores

Interactive demo of a website chat assistant: answers questions about allergens, gluten-free and vegan products and stock. The demo runs on prepared answers; the full version would use an LLM.

Tech: HTML · JavaScript

References

Translated from Danish

"We have benefited greatly from Halyna's experience with Oracle and SQL development, data analysis and system architecture. Halyna works in a structured and efficient way, with a solid understanding of relational models and periodic datasets. […] I therefore give Halyna my warmest recommendation."

Head of Department, Breeding & Genetics, Landbrug & Fødevarer · 2026

"Halyna is quick to grasp the technical aspects of a task and has solved tasks that were part of our production deliveries. Halyna will be able to contribute heavy SQL expertise in organisations with high technical demands for rigour and quality."

Head of Section, Digital Foundations, Aarhus Municipality · 2024

Danish originals and referee contact details available on request.