Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a computer vision pipeline with yolov8 opencv and ffmpeg
Senior Full Stack and Python Engineer, Mobile RE, AI Agents, SaaS Founder
I'm Py'otr — senior full-stack engineer & SaaS founder. Built interviewaibuddy.com (live AI mock-interview SaaS) and 70+ products across 9 verticals in 6+ years.
What I build:
→ Full-stack SaaS — Nex...
About this Gig
Struggling to get a computer vision pipeline that works in production? Tired of demos that fall apart on real video, real GPUs, or real resolution?
I build CV pipelines with YOLOv8, OpenCV and FFmpeg every day. Detection, tracking, batch processing, REST inference not just a notebook.
- Shipped work (some public on github.com/pyotrmuhammad):
- ComfyUI-BatchFactory AI image batch pipeline from Google Sheets.
- TikTokBatchFactory video factory with FFmpeg + orchestration.
- KrakenSubtitle frame-accurate word-by-word karaoke subtitle burner.
- SPORTSREELZ football match highlight extractor with CV + scene detection.
- Pepite Quiz (@pepitequiz on YouTube) automated quiz video channel I built end-to-end.
- nzs-luxe.com AI video generation + auto-upload to social, for my own e-commerce store.
- Podcast Clip Pipeline viral clip extractor with speaker + scene detection.
What you get: working pipeline tuned to your input and hardware. YOLOv8 model size by latency budget. OpenCV pre/post-processing. FFmpeg for ingest, encode, clip extraction. README and runnable example.
Tell me what you need detected, classified, or extracted. NDA on request.
My Portfolio
FAQ
Q: What real computer vision projects have you actually shipped?
A: ComfyUI-BatchFactory and TikTokBatchFactory are public on my GitHub (github.com/pyotrmuhammad), plus rmbg-enhance for production background removal and KrakenSubtitle for frame-accurate subtitle burning. All built and deployed by me
Q: Can you handle real-time inference or GPU optimization?
A: Yes. YOLOv8n hits 30+ FPS on a single T4 GPU. I tune model size (n/s/m/l/x) and batch size to your latency budget. ONNX export and TensorRT optimization available on Premium tier.
Q: What deliverables do I get?
A: Working Python pipeline, README, requirements.txt, one runnable example, the trained model file, and inference scripts. Premium tier adds Dockerfile and REST API server.
Q: Do you sign NDAs?
A: Yes, on every project — standard practice. Your dataset, target use case, and code stay private.
Q: How fast can you start?
A: Within 24 hours of order confirmation. Drop me your input format and target detection — I'll send a clear plan within hours.
