Vadim Bezrukov

I build automa-flow: practical APIs for files, media, datasets and web data. They fit a schedule, n8n, Make, Clay or an agent.

Try a sample input and inspect the output before you scale. Each Actor explains its limits and what it charges.

• Download in bulk: Bulk File Downloader  fetches URL lists, page images or links from another Actor's Dataset, with a file manifest.
• Convert: Bulk FFmpeg  turns up to 100 files a run into MP4, MP3, GIF or clips.
• Then: Speech to Text  for transcripts and SRT, Video Frames  for keyframes, OCR  for scanned pages.

Google Trends  returns interest over time, regions and rising queries. Schedule a watchlist for new Breakout signals; the API starter  shows one call.

• Clean rows: Dataset Cleaner  filters, dedupes, renames and casts, no JavaScript.
• Run Python: Python Code Runner  runs your main() on Dataset rows or on a schedule, with pip.
• Call an API per row: HTTP Request Batch  for Clay, n8n or Make.

Open an Issue on the Actor page with the run ID, what you expected and a sanitized example. Leave out tokens and private data.