Freelance Computer Vision Engineer — Retail Shelf Auditing

Neuvelle AI

Remote · $18.5K / 12 weeks · Freelance · Posted 5 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

We need help making our retail shelf-auditing models materially better in 12 weeks. The weak spots are known: facing counts, promo-tag detection, and ugly in-store images taken from low angles with inconsistent lighting. The dataset exists, the client pressure exists, and the internal team needs a freelancer who can improve the model rather than explain why the problem is difficult. Good handover matters because this work will stay in production after the contract ends.

Fully remote.

Key Deliverables

List the expected deliverables for this project.

Improve shelf auditing models for facing counts, promo-tag detection, and low-angle store imagery
Audit the existing labelled dataset and identify the highest-value annotation gaps
Run experiments and document trade-offs clearly for the permanent team
Deliver model artifacts, evaluation results, and an implementation handover at the end of the engagement
Coordinate weekly with the internal product and ML lead on progress and blockers

Requirements

Technical stack needed for this mission.

4+ years of computer vision engineering in retail, industrial, or other applied image environments
PyTorch for model training and experimentation
Computer Vision experience with object detection and counting workflows
Scikit-learn for evaluation support and baseline comparisons
Docker for reproducible training and inference environments
Comfort working with noisy mobile imagery captured in real store conditions