Kostiantyn

KOSTIANTYN BAKULIN
Backend / AI Automation Developer

Email: [open contact info](look above in the "contact info" section)
GitHub: https://github.com/KostiantynBk
Portfolio:https://github.com/KostiantynBk/portfolio/blob/main/README.md ​
Location: Netherlands / Remote

PROFESSIONAL SUMMARY
Junior backend / AI automation developer with 1.5 years of startup backend experience and
practical experience building AI-powered workflow tools. Experienced with Python, external
API integrations, OpenAI Responses API, structured LLM outputs, SQLite, Playwright
automation, database logging, fallback logic, and backend business workflows. Built AI
projects involving email triage, job application automation, market-signal monitoring,
structured classification, and workflow automation.

TECHNICAL SKILLS
Languages: Python, SQL, JavaScript basics​
AI / LLM: OpenAI API, OpenAI Responses API, structured outputs, prompt design,
LLM-based classification, confidence scoring, fallback logic​
Backend & APIs: REST API concepts, HTTP, JSON, external API integrations, data
ingestion, API polling loops, workflow automation​
Databases & Storage: SQLite, local persistence, database logging, audit logs,
duplicate-prevention logic​
Automation & Integrations: Playwright, email integration, Google Sheets integration,
Telegram notifications, X/Twitter data collection​
Tools: Git, GitHub, VS Code, Postman, Pydantic, Jupyter Notebook, Ubuntu/Linux basics

EXPERIENCE
Backend Engineer — Voxum

Trading automation startup app | App Store:
https://apps.apple.com/nl/app/voxum-app/id6759362982​
09/2024 – 02/2026 | Remote
 ●​ Worked as a backend engineer on Voxum, an online startup app focused on
automated crypto trading strategy signals.
●​ Integrated external crypto market data from the Bybit API into backend workflows.
●​ Contributed to backend logic for processing market data and generating
strategy-based trading signals.
●​ Worked on a strategy-builder system where users could create custom trading
strategies using predefined tools, entry conditions, validation rules, stop-loss, and
take-profit parameters.
●​ Collaborated under an NDA; private codebase cannot be shared, but high-level
architecture and responsibilities can be discussed.

PROJECTS
AI Email Triage & Reply Automation Agent
GitHub: https://github.com/KostiantynBk/emailReply
Demo: https://youtu.be/gN160zCNbAA

Tech stack: Python, OpenAI API, email integration, database logging, Google
Sheets/Telegram integration

●​ Built an AI email automation system that classifies incoming emails, extracts key
information, and generates context-aware draft replies.
●​ Designed structured outputs for category, priority, summary, required action,
confidence score, and draft reply.
●​ Added fallback logic for uncertain, sensitive, or high-priority emails so they can be
marked for human review.
●​ Stored processed emails, generated replies, statuses, and error logs for workflow
tracking and debugging.

Application Process Agent
GitHub:https://github.com/KostiantynBk/jobAnalysis
Demo: https://youtu.be/-2XxQz76ZGk

Tech stack: Python, OpenAI Responses API, SQLite, JavaScript

●​ Built an AI-powered job application assistant that analyzes job descriptions against a
predefined resume profile.
●​ Used structured OpenAI outputs to extract requirements, identify matched and
missing skills, calculate match score, and suggest project improvements.
●​ Added English and Ukrainian support for generating tailored CV bullet points and
application notes.
●​ Stored analyzed jobs, generated materials, match results, and metadata in SQLite for
application tracking.

X Market Signal Monitor
GitHub: https://github.com/KostiantynBk/copyTrading

Tech stack: Python, Playwright, OpenAI Responses API, Pydantic, systemd, Ubuntu VPS

●​ Built an AI-powered monitoring agent that tracks selected X profiles and detects
trading-related views, position updates, and market commentary.
●​ Used Playwright with a saved browser session to poll timelines, maintain login state,
and extract post text, timestamps, links, and media metadata.
●​ Integrated structured OpenAI outputs to extract tickers, sectors, sentiment,
confidence, rationale, and supporting evidence.
●​ Persisted seen posts, extracted signals, logs, and JSON reports for auditability and
duplicate prevention.

Additional projects: see full portfolio — [Portfolio link]

Additional projects: Portfolio
(https://github.com/KostiantynBk/portfolio/blob/main/README.md )

EDUCATION
Eindhoven University of Technology

Bachelor’s in Data Science | Netherlands | 2024 – Present

Relevant coursework: Programming, Machine Learning, Fundamentals of Data Analytics,
Statistics, Data Statistics, Data Structures, Linear Algebra, Math 2, Logic and Set Theory,
Discrete Structures, Applied Number Theory and Algebra.

LANGUAGES
English — C1 | Ukrainian — Native