Diffgram Reviews

There are a lot of websites that will compare and contrast Diffgram and its features with other products in the market. I want to focus instead on the development philosophy and the software development approach of the Diffgram engineering team. Diffgram, from day one, was focused on the business need and how Diffgram can contribute to our business. A year into the engagement, they have never wavered in their commitment and have been an integral part of our business processes. They collected requirements, identified gaps, prioritized features, and executed to deliver results. One of our requirements resulted in a significant software re-factor (compound file feature). Anthony and his team were relentless, broke down the part into chunks and delivered value over some time. Diffgram also worked on Pre-Signed URL Generation for AWS and other significant enhancements to amplify our product value.

Diffgram is built as a microservices and offers way more capabilities, such as query-only attributes, a query compiler and out-of-the-box reporting, which are unique and save a lot of time.

Diffgram offers many opportunities to integrate with third-party products, APIs and be an integral part of workflows. Currently, we have built some plumbing around the Diffgram product to integrate with the processes which could be simplified.

Diffgram is the best annotation tool I've used, I love the graphical output, and the customer support is quick and responsive. Our team uses Diffgram for video and image annotations, and we are able to get great productivity from it. It is straightforward & user-friendly to handle.

No Dislikes.....................................................................

We are using this tool for video and image annotation, Diffgram is very useful for us, and we have been using it for a long time, and the results have been very positive.

I'm working on a small personal ML project and I needed to annotate a huge load of images. Diffgram gave me the ability to annotate unlimited and I had no worries about billing since it's open-source. Moreover, I can automate my annotations using simple scripts.

There's nothing particular to dislike or pin point in specific.

Go for it.

Most open-source data labeling software I've used doesn't exactly feel fully open source. Also, there are certain limitations when it comes to Labeling. I've seen no such annotation limitations with Diffgram. Makes the task of labeling my training data much simpler and faster.

I really like how easy it is to upload data. The "Sync Engine" is super flexible. I can do exactly what I want for getting my tasks done. The team is very responsive to our needs. I like how extensible it is with both AWS and GCP (We are on AWS at the moment). And for video the go to frame features are exactly what we need.

Some rough edges - it's a relatively new product. Overall the team was great at responding to the few negative items we had.

We are building novel datasets. We have a complex label scheme. Responsive Support, Great data handling.