Switch to English?
Yes
Переключитись на українську?
Так
Переключиться на русскую?
Да
Przełączyć się na polską?
Tak

Valeriia Kudriavtseva

Reliable Plus holder
Offer Valeriia work on your next project.

Ukraine Kropivnitskiy, Ukraine
15 hours 1 minute back
Available for hire available for hire
15 proposals made
age 27 years
on the service 1 month 13 days
  • chat-bot
  • Supabase
  • React/TypeScript
  • Next.js
  • lending
  • telegram bot
  • Node.js
  • Claude API

Rating

Successful projects
No data
Average rating
No data
Rating
569
Web Programming
805 place out of 6509
AI & Machine Learning
64 place out of 2899
Website Development
365 place out of 2369
Online Stores & E-commerce
128 place out of 1136

Language proficiency level

Українська Українська: fluent
English English: upper-intermediate

Skills and abilities

Portfolio


  • 8000 USD

    EMBODY — D2C storefront on Next.js with AI product import and CRM

    Website Development
    Solo full-cycle D2C e-commerce for a premium women's fashion brand — 3 integrated applications in production at embody.com.ua.

    • Storefront: Next.js 16 + Supabase, 12-table DB, variants, atomic stock decrement, wishlist sync, 21-day undo.
    • Telegram product importer: GramJS + Claude Haiku Vision auto-tags categories (fallback scraper).
    • Admin + CRM: separate application — magic-link auth, multi-color CRUD, soft-delete + cron purge.

    Launch after GPM-50 UX test based on my proprietary configuration (50 AI characters + 10 reviewers) and production audit (5 QA + 5 web developers + 3 FinOps synthetic agents). 100% payment success.
  • 1000 USD

    Ad Analyst MCP — daily digest of Meta, Pinterest, Google advertising

    AI & Machine Learning
    Custom MCP server that pulls daily performance data from Meta, Pinterest, and Google Ads — and delivers a summary table in Telegram.

    How it works: Claude calls the tool → MCP server queries the Meta Marketing API, Pinterest Ads API, Google Ads API → aggregates spend / CTR / CPA / ROAS by campaigns → formats into one Telegram table with daily delta and weekly trend.

    Use case: replaces 3 dashboards + manual export to Sheets for D2C founders running paid ads across multiple channels. Anomaly flags (CPA spike, ROAS drop) pop up automatically.

    Stack: Python, Meta/Pinterest/Google Ads APIs, Telegram Bot API.
  • 1200 USD

    Telegram MCP — import of goods from the channel to the website

    AI & Machine Learning
    MCP server that allows Claude to extract posts from a Telegram channel and convert each into a product on the website — photo, description, price, options — without manual data entry.

    How it works: Claude calls the tool with channel id and date range → MCP server reads posts via MTProto → maps media to product images, caption to description, parses price/sku → pushes structured data to storefront (Supabase / API).

    Use case: small D2C brands that post new products directly in Telegram.
  • 1000 USD

    Payment Control MCP — a digest of invoices and payments for founders

    AI & Machine Learning
    Custom MCP server that scans 4 corporate Gmail inboxes for invoices, payment confirmations, and overdue reminders.

    How it works: Claude calls the tool → MCP server pulls new emails via Gmail API → classifies (invoice / payment / reminder / spam) → delivers a morning digest in Telegram with amounts, deadlines, and direct links.

    Use case: replaces the daily ritual of opening 4 mailboxes for a solo founder. Catches missed invoices and overdue B2B payments. ~15–20 min/day saved + zero missed debts over 6 months.
  • 1000 USD

    Fal.ai MCP — image and video generation as Claude tools

    AI & Machine Learning
    Custom MCP server that exposes fal.ai endpoints (Seedream v4, Kling 2.1, Nano Banana Pro) as native tools for Claude.

    How it works: Claude calls the tool with a prompt → MCP server hits the fal.ai API → returns the image/video URL + metadata. Built-in resume-cache (skips already generated assets between runs) and ffmpeg aggregation for multi-shot Reels.

    Use case: transforms a single brand brief into a full carousel or short-form video without manual work in Higgsfield UI. On top of this operates the EverStory content pipeline — 30+ posts/month with 95% visual consistency.
  • 800 USD

    EverStory — AI content pipeline for content generation

    AI & Machine Learning
    Fully automated content pipeline for a D2C memory brand on Instagram.

    Stack: Higgsfield (Seedream v4 + Nano Banana Pro + Kling 2.1) via UI, fal.ai (Seedream / Nano Banana / Kling) for programmatic generation, ffmpeg for editing.

    Brand canon (3:2 frame, two rounded rects, Helvetica, film look) — the single source of truth. Each post undergoes 4-pass QA before publication: canon / aging-rule / artefact-source / brand + caption.

    Result: 30+ posts per month with 95% visual consistency across 9 segments.

Activity

  Latest proposals 10
Freelance project
Freelance project
Freelance project
273 USD
Freelance project
Freelance project
Freelance project
Freelance project
Freelance project
Freelance project
Freelance project