We are seeking a highly skilled developers for a short-cycle feature build, bound by NDA and strict QA expectations. This project is part of a modular rollout for our platform Vietlist.com, focused on building scalable components that pass internal QA and sandbox testing before approval.
We are engaging with freelancers for full module-based development, which includes both frontend and backend responsibilities.
Here’s how the engagement works:
- Initial Requirement Sharing:
- We will provide the core objective for a specific module.
- Freelancer Qualification & Proposal:
- The freelancer must share relevant past work or experience that matches the module's scope.
- They must then provide a brief proposal, explaining how they plan to approach the task,
- NDA Signing:
- Once the proposal is reviewed and approved, the freelancer must sign a Non-Disclosure Agreement (NDA)before moving forward.
- Access to Assets:
- Upon NDA signing, we will provide:
- Full Figma design files for the assigned module.
- Any available documentation or functional notes to help them understand the flows in depth.
- Sub-Task Planning:
- After reviewing the designs and documentation, the freelancer is responsible for:
- Breaking down the module into clear sub-tasks.
- Identifying and labeling frontend and backend responsibilities.
- Sharing this sub-task plan with us for approval before starting development.
- Development & QA Process:
- The freelancer will work on the tasks as per the approved breakdown.
- Once development is complete, they are required to:
- Provide a QA walkthrough video that shows each feature working as expected.
- Highlight how each task/sub-task has been addressed in the module.
- Engagement & Payment Platform:
• • Once all the above steps are completed and the sub-task plan is approved, we will formally engage via a contract on Guru.
Task Overview
We are initiating the development of an AI-powered business data scraping pipeline focused on Vietnamese-owned or Vietnamese-serving businesses. The system will initially extract data from Google Maps using Python-based APIs, with vendors responsible for implementing and maintaining their own scrapers. The pipeline will run on a 2.5-hour cycle, normalize and categorize the data using AI, and insert valid entries into the Vietlist database. An admin panel will provide tools for monitoring, reviewing, editing, and managing scraped data. Key Components: Scraper Engine: Python-based (e.g., Scrapy, Selenium, Puppeteer – to be proposed by vendor) AI Normalization: LLaMA 3.3 model for classification, confidence scoring, and schema mapping Admin Dashboard: React/Next.js interface for real-time review, manual corrections, and oversight API Layer: Node.js middleware for database communication Database: PostgreSQL for storing normalized and finalized records Scheduler: 2.5-hour cron job for automated scraping and processing User Roles: Admin: Full control including AI overrides and log access Reviewer: Edit and approve/reject entries, limited to content actions Read-Only: View-only access System Bot: Automates scraping, inserts entries, logs actions Important Note: For the initial rollout, only Google Maps will be scraped. Vendors are expected to develop and maintain their scraping logic, while the rest of the pipeline will be built by client.