A pioneer in using AI for natural language processing (NLP), machine vision, and speech recognition, Spell has released the world’s first cloud-agnostic, end-to-end deep learning platform. With the company’s namesake solution – designed by AI industry veterans – users can track, manage, and automate the entire deep learning (DL) workflow from model creation to deployment and optimization. In both large enterprises and AI startups, the platform dramatically improves efficiency, effectiveness, and compliance. Spell offers the most comprehensive MLOps platform for deep learning.

Spell’s CEO and co-founder, Serkan Piantino, commented, “The rapid growth of the Spell user community is a gratifying validation of our vision for democratizing deep learning through a comprehensive, transparent platform for enabling and accelerating the successful adoption of advanced AI across a broad range of industries and use cases, and we are just getting started.”

MLOps solutions are typically designed for machine learning with traditional processes and infrastructure. They use commodity CPU computing resources to build, train, and deploy predictive models. In a deep learning context, hundreds of experiments with thousands of parameters can be managed at once, requiring users to track and manage a vast amount of more costly computation. Without effective methods for managing deep learning (DLOps), organizations struggle to implement deep learning applications such as NLP, machine vision, voice recognition, and others, or they can only do so with a high cost of implementation.

Through Spell, deep learning models can be orchestrated, documented, optimized, deployed, and monitored throughout their entire lifecycle. The ad hoc point tools for each of these functions do not provide an integrated view of all the steps involved in the end-to-end development and deployment process.

The Spell platform captures and tracks detailed information about every aspect of the model life cycle, such as data, parameters, experiments, personnel, computing resources, infrastructure, deployment location, and more. As a result, a deep learning project is accountable, collaborative, and productive, and it can be reproduced, explicated, and governed throughout its lifespan.