Contact information

You can get this candidate's contact information from https://www.work.ua/resumes/20927329/

Work experience

Freelance Python Backend Engineer

from 09.2026 to now (3 months)
Supplax (freelance), Київ (IT)

Competitor data collection pipeline (freelance, remote)
• Problem: The marketing team spent 15+ hours per week collecting competitor data from Yelp (business profiles, ratings, reviews) by hand: 6 competitors in each of 15 cities.
• Solution: Built a Python web scraping pipeline that tracks 90 competitors in 15 cities and separately collects new business listings on Yelp, with scheduled runs, request throttling, retry logic and error handling; cleaned, deduplicated and validated the data with Pandas.
• Result: Manual research eliminated (15+ hours per week saved). Marketing receives structured, analysis-ready datasets on schedule.

Stack: Python, Pandas, web scraping, ETL, scheduled jobs.

AI & Automation Engineer (Python Backend)

from 06.2025 to 05.2026 (1 year)
HolyWater Tech, Київ (IT)

1) Fully automated video localization
• Problem: Each motion designer localized 10-15 video creatives per day by hand, every creative needed 6 language versions, and the plan was to scale to about 80 creatives per day. Manual work caused errors: wrong or literal translations on text overlays.
• Solution: Built 3 Python bots that repeat the full workflow: take a task from Airtable, open the project in DaVinci Resolve, translate on-screen text with LLMs, run a pre-render technical check, render, run an AI visual QA and upload to Google Drive. LangChain routes work between a cheaper text model and a separate vision model.
• Result: Replaced the manual workflow of a 3-person team, volume grew 4x with almost zero errors, no human involvement.

2) Marketing (UA) automation: automated creative launches via ad platform APIs and daily ad metrics reporting (spend, CTR, installs, CPI); UA managers stopped manual uploads and reporting.

3) Creative Builder: Premiere Pro plugin (JavaScript, ExtendScript) that assembles new ads from tested elements; production cycle -70%, about 1 in 5 such creatives became top-performing.

4) Creative performance analytics: database of ad elements to find what works in top ads.

Stack: Python, Django, FastAPI, PostgreSQL, MongoDB, Redis, Docker, AWS (EC2, S3, Lambda), GitLab CI/CD, OpenAI and Claude APIs, LangChain, n8n, Airtable API, Google APIs.

Motion designer

from 05.2024 to 06.2025 (1 year 2 months)
HolyWater Tech, Київ (IT)

Ad format adaptation scripts
• Problem: Every video ad had to be re-cut by hand into several formats for different ad platforms (Google Ads / YouTube, Meta, TikTok), which took a large part of the motion team's time.
• Solution: Wrote about 20 automation scripts for the design workflow: resizing creatives to 16:9, 1:1, 4:5 and 9:16, keeping text and key elements inside platform safe zones, batch export with platform presets and file naming, and replacing text overlays from a template.
• Result: Format adaptation became an automated step instead of manual rework, so designers spent more time on new creatives. This work led to a move into a full-time engineering role.

Also produced motion graphics and video ads for marketing campaigns across several product lines.

Graphic designer

from 02.2024 to 04.2024 (3 months)
Welltech (freelance), Дистанційно (IT)

Created digital assets and visual communications for health and wellness products (remote, freelance).

Education

Київський національний університет технологій та дизайну (КНУТД)

Design and Visual Communications (Bachelor), Київ
Unfinished higher, from 2023 to 2027 (4 years)

Knowledge and skills

• Бажання вчитися і розвиватися
• Python
• Git
• SQL
• GitHub
• PostgreSQL
• Docker
• Django
• FastAPI
• REST API
• Claude API
• Google API
• LLM API
• Telegram API
• AWS
• Docker Compose
• CI/CD
• Адаптивність
• Знання принципів ООП

Language proficiencies

• English — average
• Ukrainian — fluent

Additional information

Python Backend Engineer with nearly 3 years of commercial experience in tech (since 2024) and 2+ years of building automation with Python. I turn manual, repetitive work into reliable automated systems: at HolyWater Tech I built AI-powered bots that took over video localization end to end (6 languages, 4x volume, near-zero errors, no human in the loop) and marketing automation for the UA team. I build backend services with Django, FastAPI, PostgreSQL, Redis, Docker and AWS, and integrate LLMs (OpenAI, Claude, LangChain) into production workflows. My background in motion design helps me understand the processes I automate.

Open to: full-time or part-time, remote or hybrid in Kyiv. Ready to start within 1-2 days.
GitHub: github.com/lessiatk
LinkedIn: [open contact info](look above in the "contact info" section)