Backend Development for Multi-Platform Marketplace Alert

Job ID: 40548853

Budget: $30 – $250 USD

Title: Build a Real-Time Marketplace Alert System (Facebook, OfferUp, eBay, and should be scalable)

---

OVERVIEW

I need a developer to build a backend system that monitors Facebook Marketplace, OfferUp, and eBay for new listings in any city a user selects, then sends instant push notifications when new listings are found.

I already have a working dashboard (Node.js) and mobile app connected to it. You are NOT touching the frontend. You are building a scraping/alert backend that plugs into my existing system via a simple REST API endpoint.

---

WHAT YOU WILL BUILD

1. A scraper that monitors Facebook Marketplace, OfferUp, and eBay for new listings by city and keyword
2. A deduplication system that tracks already-seen listing IDs so users only get alerted to genuinely new listings
3. A REST API endpoint that my existing dashboard can poll to get new listings (same JSON format as my current Apify setup)
4. Push notification delivery via Firebase Cloud Messaging (FCM) when new listings are found
5. Deployed and running on Railway or Render (I will provide accounts)

---

TECHNICAL REQUIREMENTS

Facebook Marketplace:
- Must use warmed Facebook accounts + ISP proxies (I will purchase proxies from Webshare)
- Must use Playwright to manage sessions and intercept GraphQL responses
- Must rotate between multiple account cookie files if one gets blocked
- Must filter by city and keyword passed as parameters
- Poll interval: every 2–5 minutes

OfferUp:
- Lightweight Playwright scraper
- Filter by keyword, location, and price range
- Poll every 2–3 minutes

eBay:
- Use eBay's native RSS feed for new listings (free, no proxies needed)
- Filter by keyword, location, and price range
- Poll every 60 seconds
- Sort by newly listed to catch fresh listings immediately

Output format for ALL platforms (must match this exact JSON structure):
{
"id": "listing_id",
"title": "Item title",
"price": 150,
"location": "Toronto, Ontario",
"url": "https://...",
"image": "https://...",
"platform": "facebook" | "offerup" | "ebay",
"listed_at": "2026-06-30T12:00:00Z"
}

---

WHAT I PROVIDE

- Access to my existing Node.js dashboard codebase (GitHub)
- Webshare ISP proxy credentials
- Firebase project credentials for FCM
- Railway/Render account for deployment
- 5 Facebook accounts (I will warm them before handing over)
- Clear documentation of my existing API format

---

WHAT I DO NOT WANT

- No Apify (too expensive at scale)
- No per-result pricing APIs
- No touching the frontend dashboard or mobile app
- No solutions that require me to be logged into a browser manually each time
- No solutions that break after 1 week with no support

---

DELIVERABLES

1. Working backend deployed on Railway/Render
2. All source code pushed to a private GitHub repo I own
3. README with setup instructions including how to add new FB account cookies and rotate proxies
4. All 3 platforms tested and confirmed working before handover
5. 2 weeks of bug fixes after delivery included

---

IDEAL CANDIDATE

- Has built web scrapers before, specifically for Facebook or other protected platforms
- Knows how to handle anti-bot detection (ISP proxies, session management, Playwright stealth)
- Experience with Node.js or Python backends
- Can communicate clearly in English
- Has verifiable past scraping projects in their portfolio

---

TO APPLY

Please answer these 3 questions in your proposal, or I will not respond:
1. Have you scraped Facebook Marketplace or a similarly protected platform before? What approach did you use?
2. How will you handle Facebook blocking the scraper? What is your fallback strategy?
3. What is your realistic timeline to deliver all 3 platforms working and deployed?

Proposals that say only "I can do this" with no answers will be ignored.

---

This is a straightforward project for an experienced scraping developer. The core logic is well-understood — I have done the research. I need execution, not discovery.