AI Product Manager who ships 0→1 products from an ambiguous problem to a working release.

I combine product discovery, specifications, rapid prototyping, analytics, QA, and go-to-market execution. I use AI tools and code when they shorten the path to real user feedback.

LinkedIn · Email · Kyiv, Ukraine · Remote

A serverless publishing pipeline using the official Meta APIs and GitHub Actions. It renders content, publishes on a schedule, prevents duplicates, and commits durable state back to Git.

• Repository history records 121 Threads posts, 90 Instagram carousels, and 209 automated publishing commits.
• Runs without a server or an always-on computer.
• Keeps tokens in GitHub Actions Secrets and serializes jobs with concurrency controls.

A browser-based sales-training and product-demo tool. Give it a public Instagram profile or website, review the extracted business persona, and role-play a realistic DM conversation with an AI setter.

• Shipped from idea to a live Next.js product in 5 days.
• Designed the full flow: source analysis → editable persona → grounded conversation.
• Replaced Meta Graph API dependence with an Apify-based path to reduce onboarding friction.
• Added explicit anti-hallucination rules and credential gating to control API cost.
• Stack: Next.js 15, React 19, TypeScript, OpenAI API, Apify, Tavily, Vercel.

A multilingual acquisition and qualification funnel for an AI appointment-setting product.

• Shipped 3 offer generations across English and Dutch.
• Built A/B infrastructure for concurrent variants, copy, form length, and urgency tests.
• Connected lead capture, Calendly scheduling, and server-side Meta conversion tracking.
• Implemented event-ID deduplication and hashed PII for cleaner attribution.

A responsive real-estate discovery experience covering property search, listing pages, content, and lead capture.

• Reworked the information architecture and mobile-first interface.
• Added responsive WebP assets, hero preload, lazy loading, and lighter scroll behavior.
• Reduced built CSS from roughly 300 KB to 118 KB before deployment.
• Validated the production build and shipped the optimized version to a public URL.

• Took an AI B2B product from idea to production in 2 months, owning feature definition, acceptance criteria, QA, onboarding, and release decisions.
• Specified and shipped an internal quoting product with one developer: 15 daily users, 3 releases in 25 days, and a 12-page branded PDF generated from margin, commission, and VAT rules.
• Delivered 50+ websites and landing pages end to end, including discovery, scope, launch, SEO, analytics, and conversion tracking.
• Built data and content workflows covering 23,431 qualified business records and 209 automated publishing commits.

1. Name the user, the painful problem, and why it matters now.
2. Define the smallest test that can disprove the idea.
3. Ship a clickable artifact in hours and a usable v1 in days.
4. Instrument the critical behavior, watch real usage, and iterate.
5. Document decisions and hand over a clear production path.

Product: discovery, PRDs, user stories, acceptance criteria, roadmaps, prioritization, experimentation, pricing, QA
Build: Python, TypeScript, Next.js, SQL, Playwright, REST APIs, GitHub Actions
AI: Codex, Claude Code, Cursor, LLM workflows, retrieval, evaluations, prompt systems
Analytics and growth: PostHog, GA4, Meta CAPI, funnels, landing/VSL experiments

More detail: AI Product Builder Portfolio