AI + IoT

As computers become ubiquitous in the environment, AI and the Internet of Things are becoming ever more important as new places to do business.

Want faster model development? Get the "2019 State of Development and Operations of AI" report now.

Get the report

With the exponentially increasing amount of data coming in real time from sensors in all kinds of environments in the world, AI and IoT on the Edge is becoming a crucial use case for more and more businesses. AI is already a challenge on its own, with big data, machine learning, specialized expertise, regulatory requirements, and deployment to production.

IoT is another challenge, with sensors, huge real-time data, specialized hardware, and new use cases. Many companies are already struggling with just one of these two challenges. Combining the two is even more difficult, yet the combination of ubiquitous data plus AI to create value from it offers huge opportunities for revenue for those businesses that can realize its value.

Key AI challenges

  • AI and IoT on the Edge needs to deal with massive streaming data, often on specialized hardware of limited size, which precludes many turnkey solutions.
  • But the requirements of enterprise machine learning nevertheless remain: reproducibility, accountability, collaboration, and continuous delivery.
  • Combining these requirements only adds to the challenge, making it so difficult as to be intractable for many businesses. Yet it is simultaneously becoming crucial for them to address.
How can Dotscience help?

Dotscience has many strengths including our ability to transfer large amounts of data through a pipeline without having to make copies (leveraging our core ZFS architecture), our ability to do compute on almost any machine (bring your own runner), and our easy deployment of models as containerized microservices including onto Kubernetes.

This means that we can deal with large datasets for training of models, and then deploy them to do real time inferencing on the Edge in places that are really needed to solve AI+IoT business use cases.

Combine this with other core ideas from our product — reproducibility, data provenance, auditability, collaboration, plus an underlying architecture developed by a team of DevOps and machine learning experts — and you have a product with the strength of foundation and flexibility of set up to make AI on the Edge work for real projects.