TikTok Trends Parser
2026Full Cycleinner-tools
Role
Full cycle: from idea and architecture to development and deployment.
Tech
| Layer | Stack |
|---|---|
| Backend | FastAPI, SQLAlchemy, Alembic, Celery + Beat |
| Frontend | Next.js 14, TypeScript, Tailwind CSS, Zustand |
| Data | PostgreSQL 15, Redis 7 |
| Parsing | Apify API (TikTok scraper) |
| Auth | JWT + Telegram Bot (login via Telegram, no passwords) |
| Deploy | Docker Compose (dev + prod), automatic migrations |
Key features
- 01Scheduled automatic parsing. The system collects videos by configured hashtags — weekly, without human involvement. Each run is logged: how many found, how many new, how long it took.
- 02Smart hashtag prioritization. The algorithm decides which hashtags to parse first: promotes "hot" ones (many new videos), penalizes "stable" ones (few changes), speeds up those unchecked for a while. Not a FIFO queue — a scoring model based on parse history.
- 03Team collaboration with tracking. Each user sees their own bookmarked, viewed, and used videos — without interfering with others. Schemes can be shared between team members with access control.
- 04Filtering and export. Flexible filters: by date, views, likes, engagement rate, author, hashtag. Results exported to Excel in one click.
- 05Webhook integrations. Daily automatic data export to external systems (Zetter, SHARE, and any others) — the parser acts as a data source for your pipeline.
01
Problem
Manual TikTok monitoring is a routine that doesn't scale. Content managers spend hours scrolling through hashtags, miss trends, can't react to hype in time, and results can't be systematized or shared with colleagues. When you're tracking dozens of hashtags across multiple content schemes — the manual approach simply breaks down.
02
Audience
| Role | Scenario |
|---|---|
| Content manager | Reviews new trending videos in their niches daily, bookmarks favorites, exports a selection for production |
| SMM strategist | Creates hashtag schemes for different brand directions, analyzes engagement rate, identifies content patterns |
| Content department head | Sees team activity, manages access, monitors trend coverage via admin panel |
| Analyst / researcher | Collects data on specific hashtags with metrics for reports and market research |
| Agency | Manages multiple clients through separate schemes with result sharing |
03
What's different
vs. manual monitoring:
- Parsing runs 24/7 on schedule — trends aren't lost between sessions.
- Metrics are collected automatically — no manual note-taking.
- Full history saved and available for filtering.
vs. typical parsers:
- Not just "download videos by hashtag." A system with users, roles, personal status tracking, and parse history.
- Smart prioritization — the parser adapts to each hashtag's dynamics, not a dumb list iteration.
- Webhook export — the parser integrates into existing workflows, not isolated.
vs. SaaS platforms:
- Full data control — everything stored on your server.
- No subscription per hashtag or user count.
- Unlimited customization — the code is yours.
04
Screenshots
05
Customization
- 01Parse schedule — day of week, hour, minute (env variables).
- 02Limits — max videos per run, results per hashtag.
- 03Hashtags — weight (1-4), minimum views, priority order — all via UI.
- 04Webhook endpoints — connect any external systems for automatic export.
- 05Roles and access — admin/user, scheme sharing between users.
- 06Deploy — ready Docker Compose configs for dev and production, all settings via `.env`.
Need something like this?
Tell me about the task — I'll suggest an approach, pick a stack and scope it out.