VendorForge AI: An Autonomous Reseller Copilot That Turns Clutter into Cash

Описание и применение

VendorForge AI автоматизирует исследование, ценообразование и создание описаний для продавцов на платформах Etsy, eBay и Shopify, экономя более 3 часов работы и устраняя расходы на токены, при этом защищая конфиденциальные данные.

Задача

Resellers face a “Death Pile” time sink of 15‑30 minutes per item for research, SEO, and multi‑platform listings, and suffer from pricing guesswork that can erode margins or tie up inventory, while closed‑API SaaS tools add 10% token‑fee costs and risk exposing trade secrets.

Результат

Resellers reported saving over 3 hours of listing labor per Sunday batch and confirmed willingness to pay $15/month, while open‑weight models eliminated token fees and protected trade secrets.

Рабочий процесс

• 1. Intake sourced item and photo specs
• 2. Query MongoDB Atlas Vector Search for semantic memory
• 3. Scrape live eBay & Etsy sold comps via SerpApi
• 4. Run TabPFN‑3.5 for price, sale probability, and anomaly flagging
• 5. Generate platform‑specific copy with Google Gemma 2
• 6. Produce 15‑second voiceover with ElevenLabs
• 7. Record telemetry with Sentry Agent Tracing

Доказательства

After sourcing 20 items at a thrift store or estate sale, a seller spends 15 to 30 minutes per item manually researching sold comps, writing SEO titles, counting tags, and tailoring descriptions for 3 different platforms.

Источник

In our validated pilot, a reseller managing 2,400+ online sales confirmed a willingness to pay $15/month after saving over 3 hours of listing labor on a single Sunday batch.

Источник

Running Gemma 2 open weights locally or on shared instances brings marginal inference cost to $0.00.

Источник