Social Media Scraper for Restaurant UGC Video Project

We are looking for a contractor to help us build a scalable scraping system for collecting restaurant-related UGC videos from social media. UGC means User Generated Content. For this project, we are looking for videos created by real customers, food reviewers, vloggers, and creators who visit restaurants and post about their experience. The first focus city is Birmingham, UK, but the system should be designed so it can later work across multiple cities. Project Overview We are building a long-term pipeline that can discover, collect, and organise restaurant UGC videos, then prepare them for matching against restaurant records in our database. The scraper’s job is not just to collect one batch of videos. We want a repeatable discovery system that can improve over time, reduce noise, find new accounts and videos automatically, and eventually scale into other cities. Videos may later be matched to restaurants using evidence such as captions, tagged locations, tagged accounts, audio mentions, signage, branding, menus, packaging, interiors, or addresses shown in the video. Accuracy matters more than volume. We want useful videos with minimal noise. What You’ll Build We need help creating 3 scraping/discovery systems: 1. Food Page Discovery System Find and organise Birmingham, UK food pages, reviewers, creators, local guides, and accounts that regularly post restaurant UGC. Deliverables: List of relevant food/UGC accounts Account URLs, usernames, platform, and account type Notes on why each account is relevant Repeatable method for finding more accounts Logic that can later be reused in other cities 2. Restaurant Video Discovery System Find restaurant-related videos using search, APIs, and scraping methods. This may include: Keyword searches Hashtag searches Location-based searches Apify, RapidAPI, or similar tools Puppeteer, Playwright, or other non-API scraping methods Discovery through related accounts, similar videos, and creator networks Deliverables: Video URL Platform Creator username Caption, if available Tagged location/account, if available Search source used Basic relevance score or notes 3. Video Collection & Review Pipeline Create a clean process for storing, filtering, deduplicating, and preparing videos for review. Deliverables: Structured dataset of collected videos Deduplication process Status field: pending, approved, rejected, unsure Notes explaining why a video is relevant or rejected Handoff format for the extraction/matching team Feedback loop so rejected/approved videos can improve future scraping logic Working Requirements This is a 2-week project. The contractor must be available to work within a team that includes data scientists, extraction specialists, and project leads. You must be willing to: Work weekends if required Join daily stand-up meetings at 10:00am UK time Communicate blockers clearly Share progress regularly Work quickly while keeping data quality high Adapt the scraping approach based on feedback from the wider team Ideal Candidate You should have experience with: TikTok, Instagram, or social media scraping Apify, RapidAPI, Puppeteer, Playwright, or similar tools Scraping around keywords, hashtags, locations, captions, bios, and accounts Producing clean structured datasets Reducing noisy or irrelevant results Building repeatable and scalable scraping workflows Designing systems that can improve from feedback over time Bonus if you have worked with food, restaurant, local business, location-based, or creator datasets before.