develop your own website

We accelerate AI applications

The Data Management solution that redefines how large visual data sets are stored, searched and processed.

      We believe that AI requires a new type of data management


A solution enabling Data scientists and AI developers to handle large data objects without the burden of juggling across multiple disconnected systems 

Easy to use

A simple interface that allows applications to easily store and extract insights from visual data. 

AI ready

 A solution that saves on precious compute cycles by natively supporting pre-processing steps commonly required by visual AI or ML pipelines.

Efficient and evolving

 A solution that scales and allows data and its associated metadata to grow seamlessly as AI models improve. 

ApertureData platform embodies our core beliefs

Built on the VDMS (Visual Data Management System), ApertureData Platform makes deployments of AI applications more easily achievable and seamlessly scalable.

Unify metadata and data

Through an intuitive JSON API that unite metadata search, data retrieval and preprocessing operations.

Allow metadata to evolve easily

By storing metadata in a hardware optimized graph database we allow efficient search and seamless evolution.  


Acknowledge visual data

Through data formats, API primitives and operations optimized for retrieval and preprocessing.

Host anywhere

ApertureData Platform is offered on premise or as a cloud service to fit your needs.

Try open source first

Our core VDMS source code is available on github and we have a docker option. VDMS server can listen to multiple clients. We have a C++ and python client library.  Learn more...

Write JSON commands

We define simple JSON commands to expose a unified interface for metadata and data avoiding any need to worry about the underlying storage solution.  We also recognize visual entities like images, videos, bounding boxes, and feature vectors in our API. Learn more...

Start managing data

We support multi-modal, easily evolving metadata store and provide traditional as well as novel formats for data. The metadata can extend beyond just information for visual entities e.g. information about people, events etc, while the data other than visual can be stored as blobs. Learn more...

Build your ML pipelines

Using the JSON commands, you can find the relevant subset of data, pre-process it and feed it to a machine learning pipeline or answer data science queries. You can then go ahead and update the existing data or metadata to keep enriching your insights repository. Learn more...


Vishakha Gupta

Founder and CEO
Vishakha has over 7 years of experience at Intel Labs in designing and developing various systems technologies optimized for new hardware after a PhD in Computer Science from Georgia Tech and MS in Information Networking from Carnegie Mellon University. 

Marc Dupaquier

Strategic Advisor
Marc is Managing Partner at Novaroutes, a Business Angels partnership specialized in B2B Artificial Intelligence Software. He is also Managing Director at Cartesiam, one of NovaRoutes investments and the Global Leader of AI at the Edge.

Prior to this role, he led multi-billion dollar businesses at IBM, including the Business Partners Group, the iconic Mid-Market Division and Global Marketing for the Hardware and Software Groups. 


We would love to chat and learn how we can help you or work with you.