Amar Krishna

Matrice.ai - Integrating AI models into your apps has never been easier

by
End to end ML platform from annotating data sets to model training (using our machine learning/ deep learning model library) to model inference to model deployment. Trained model then can be exposed to the end user using APIs or choice of their cloud service.

Add a comment

Replies

Best
Amar Krishna
Hello folks on PH, this is Amar, co-founder of Matrice.ai. I previously built Chefling and the experience of annotating datasets, training models and deploying them on then edge was a painful experience. We wasted months of resources without any real outcome. The whole experience led us to launch Matrice.ai. We are three co-founder with an extensive background in machine learning. The Matrice.ai platform provides the end user with an environment to swiftly upload their data to the platform; both labeled and unlabeled datasets can be uploaded. In future iterations, the user will have tools to enable them label images by selecting one or more of the user’s pre-defined labels. Also, the end user has the ability to add bounding box; to be used for object detection and segmentation tasks. The user can then utilize a plethora of deep learning models available at their finger tips. Once a pre-trained model is selected from the model library, the user can train the model on their dataset to fine tune the model to their dataset. A trained model then can be deployed and exposed to the end user where they can query the model for inference on new data points.