Личные данные скрыты

Yuliia Shkvarchuk
Product Manager | Data Quality Manager
Ivano-Frankivsk, Ukraine | [открыть контакты](см. выше в блоке «контактная информация») | Email | LinkedIn​

Specialist in Knowledge Management and Data Quality with a career built on a deep obsession with data integrity and structural
precision. Expert in translating complex backend logic and regulatory requirements into clear, user-centric documentation. Proven
track record in managing zero-defect localization releases and building structured knowledge systems for AI-automated workflows.

AREAS OF EXPERTISE
Product & Market Strategy: Market Sizing & Research | Competitor & Traffic Analysis | Market Entry Barriers | Customer Discovery
Leadership & Delivery: Cross-functional Leadership | Agile/Scrum Delivery | Backlog Prioritization | Requirement Gathering (PRD)
Technical & Growth: AI Agent Orchestration | System Architecture | SQL Data Analysis | SEO Growth

TECHNICAL SKILLS
AI & Automation: OpenAI Assistants/Tools | n8n | Flowise | VAPI | Langfuse | Webhooks | Prompt Engineering, including
multi-section system prompts with behavioral specifications, safety protocols, anti-hallucination rules, and structured output
enforcement
LLM Operations: Model Selection & Optimization | Token Cost Management | Context Caching | Gemini & Anthropic APIs
Data & QA: AI Test Case Development | SQL (Metabase) | Debugging (Browser/Server)

PROFESSIONAL EXPERIENCE
Inventory Product Manager | Jome Oct 2024 - Present

Inventory Product Manager focused on data integrity and system architecture for a high-volume real estate marketplace. I own the
quality surface of the inventory product, implementing AI-driven auditing and structured internal tools to ensure 100% data accuracy
across automated workflows.

●​ Prompt Architecture & Rigor: Developed a multi-section prompt framework with explicit behavioral specifications and
anti-hallucination rules. Iterated based on live-call transcript audits to fix failure modes, prioritizing behavior fidelity and
technical reliability over simple automation.
●​ Audit-Driven Optimization: Built 100+ Metabase queries to provide real-time visibility into inventory health. These
monitoring tools directly reduced manual data quality efforts by 30% through automation of systematic error identification.
●​ Data Architecture: Designed a multi-source merging logic to de-duplicate listings from scrapers and MLS feeds.
Established priority hierarchies and automated image imports, allowing the system to scale without increasing the Data
Quality team size.
●​ Structural Data Engineering: Led the transition from a flat listing model to a tiered architecture (Parent Community →
Collection). Personally oversaw the structural reclassification of 3,496 communities (30% of inventory) to ensure
architectural precision and prevent data fragmentation.
●​ Technical Documentation: Translated complex backend logic and multi-source merging rules into clear internal
documentation, ensuring the Data Quality team could scale without increasing headcount.

PROJECT SPOTLIGHTS

Alex AI — SMS & Voice Lead Engagement System

●​ The System: Designed ground-up behavioral logic for a Vapi.ai voice/SMS agent. Built an "Outcome
Disambiguation Framework" to define precise linguistic rules for distinguishing "not right now" (future prospect)
from "genuinely done" (lost).
●​ Rigorous QA: Built dedicated test suites to validate agent behavior across complex edge cases and ran daily
transcript audits to iteratively refine system logic.
●​ The Outcome: Eliminated intent misclassification and reactivated a cold pipeline, resulting in a 3.47%
contact-to-booked rate across 25,000+ interactions.
Low-Cost SEO Automation (Jome)

●​ The Challenge: A requirement for 2,000+ unique landing pages while maintaining zero-defect quality at a low
per-unit cost.
●​ The System: Designed a multi-section prompt architecture that enforced strict behavioral rules and
anti-hallucination guardrails.
●​ The Outcome: Improved search visibility by 6 positions site-wide in 4 months, achieving 100% coverage for under
$60—replacing high-cost manual agencies with a structured, automated system.

Platform Data Architecture (Collections)

●​ The Challenge: A legacy flat architecture.
●​ The System: Defined and closed gaps in multi-source merging logic, including automated image imports,
market-level merge blocking, and creation-date validation.
●​ The Outcome: Established a self-sustaining deduplication system that handles thousands of new listings
per week without requiring additional QA headcount.

OTHER JOB EXPERIENCE
Data Labeling Specialist | NDA Feb 2024 - Jul 2024

Worked on the development of AI through accurate and detailed data labeling.

●​ Visual Data Annotation: Authored detailed English descriptions, executed targeted media recording tasks, and performed
precise object-level segmentation to deliver high-quality training inputs for computer vision pipelines.
●​ Dataset Curation & QA: Researched, reviewed, and filtered complex visual assets to organize them into structured,
high-purity collections while strictly adhering to rigorous labeling guidelines.
●​ Compliance & Collaboration: Maintained strict data privacy and confidentiality protocols while facilitating clear
communication with international stakeholders to align on project criteria and volume targets.

Data Quality Manager & Inventory QA | Jome May 2022 - Oct 2024

Evolved from managing large volumes of data to effectively auditing system health through the use of statistical analysis.

●​ Inventory Integrity: Built and maintained efficient inventory tracking systems for new construction data, focusing on
long-term data accuracy and structural consistency.
●​ Process Optimization: Collaborated with content and IT departments to optimize inventory-related processing methods,
reducing systematic errors in the data pipeline.
●​ Builder Relations: Established structured communication channels with builders to ensure high-quality, up-to-date content
for the platform.​

Subtitle & Dubbing Editor, QC | MEGOGO Sep 2020 - May 2022

Core Focus: Managing zero-defect localization releases for a top-tier streaming platform through rigorous quality control and asset
management.

●​ Zero-Defect Release Management: Oversaw end-to-end technical QC for digital video/audio assets, maintaining 100%
compliance with international localization standards for high-profile global titles.
●​ Asset Database Integrity: Managed complex asset databases using Google Sheets and Excel to track project dynamics
and ensure timely, accurate content releases across diverse clients.
●​ Technical Audit & Script Review: Evaluated translation quality and technical subtitle synchronization, solving localization
challenges through systematic script reviews and cross-functional coordination.​

EDUCATION AND CERTIFICATION
Master of electronic multimedia technologies | Chernivtsi National University (2014-2015)
Bachelor's in Publishing and Printing | Chernivtsi National University (2010-2014)
Google Data Analytics: Mastered SQL, Python, and Tableau for end-to-end data cleansing, analysis, and visualization to drive
strategic decision-making.

Product Management: Building AI-Powered Products: Developed skills in AI agent orchestration, LLM prompt optimization, and
managing data pipelines for production-grade AI features.

Generative AI: Prompt Engineering: Specialized in chain-of-thought (CoT) prompting and iterative debugging to generate precise,
reliable outputs from Large Language Models.

OSINT and Information Intelligence: Mastered data collection and verification techniques using open-source tools; gained
expertise in investigative methodologies, information security, and public data source analysis for national security contexts.