Keymakr Data Labeling

BACKGROUND

Introduce your business and what you do there.

I’m the tech lead at Floware, an IT company. We provide advanced mobility analysis and prediction solutions using integrated sensors, modeling tools, and AI to support decision-makers in managing transportation flows.

OPPORTUNITY / CHALLENGE

What challenge were you trying to address with Keymakr Data Labeling?

We needed help with providing annotated datasets for detection models.

SOLUTION

What was the scope of their involvement?

Keymakr Data Labeling sourced and annotated data. The team sourced videos and photos and annotated them according to preset classes. The team delivered multiple datasets in COCO format for YOLO usage.

What is the team composition?

We worked with an unknown number of teammates from Keymakr Data Labeling.

How did you come to work with Keymakr Data Labeling?

We found them through an online search. We chose them because they offered good value for cost and services.

What is the status of this engagement?

We worked together from June to December 2025.

RESULTS & FEEDBACK

What evidence can you share that demonstrates the impact of the engagement?

We managed to train powerful models from the provided datasets. Our detection rate improved from 10%–80% for specific classes in our dataset.

How did Keymakr Data Labeling perform from a project management standpoint?

Keymakr Data Labeling was very helpful and fast. The team delivered on time despite a tight schedule. We communicated via in-person meetings, emails, and messaging apps.

What did you find most impressive about them?

The constant feedback loop and quality of communication were impressive. It helped shape the project on the go to tackle challenges.

Are there any areas they could improve?

There’s nothing they can improve on.

BACKGROUND

Introduce your business and what you do there.

I’m a data engineer at a manufacturing company. We design and manufacture high-quality audio products, including microphones, networked audio systems, signal processors, and headphones for professional and consumer use.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Keymakr Data Labeling to accomplish for your data engineering work at Shure?

We needed to make training our data easier and more reliable.

SOLUTION

Can you describe the scope of work Keymakr Data Labeling handled for this project?

Keymakr Data Labeling provides bounding box annotations for faces and bodies of people in videos and images. We give them the videos and images to annotate, and they handle everything after that. The team takes our feedback and provides us with JSON files with the bounding boxes, confidence attributes, and reflections according to our needs.

What is the team composition?

I'm not sure of the exact number, but they always provide us with a team to fit our needs and timeline expectations

How did you come to work with Keymakr Data Labeling?

We chose them over other options because they were a great culture fit, offered good value for the cost, and had company values aligned with ours.

What is the status of this engagement?

We started working with them in May 2023, and our engagement is ongoing. 

RESULTS & FEEDBACK

What evidence can you share that demonstrates the impact of the engagement?

Keymakr Data Labeling is very on time and never late. They always take our feedback and respond very promptly, even though there's quite a lot of time difference. Their progress and success have been shown in how they've improved with the feedback and are willing to take it and iterate as many times as possible. Overall, we've noticed more quality annotations with fewer errors, better consistency, and better-formatted annotations to fit our needs.

How did Keymakr Data Labeling perform from a project management standpoint?

Keymakr Data Labeling's project management is amazing. The team is very nice and patient and responds very quickly. They tell us the time and deliver everything on time. We communicate via in-person meetings and email or messaging apps.

What did you find most impressive about them?

Keymakr Data Labeling's openness to iterations and feedback is impressive.

Are there any areas they could improve?

There's nothing they could improve on; they're perfect.