2020: A year in review
It seems almost trite to now say that 2020 has been a year like no other. It’s been a year of rolling with the punches, with each of us still learning to manage and adapt in the best way we know how. COVID alone has forced organizations to do away with well thought-out product roadmaps and rethink survival. Instead of keeping up with emerging technologies, we find ourselves asking, “What do we need in order to survive, and what exactly can we afford NOT to do?”
Highly experimental projects have given way to more promising bets, compressing what would have otherwise been years of experimentation down to months. And market dynamics have forced leaders to re-evaluate trade-offs between near term and long-term investments with a new sense of urgency.
91% of 520 global executives surveyed by the Harvard Business Review said that their organizations changed their operating models over the past ten months, confirming that a successful digital transformation requires innovation at both enterprise and at business process levels.
The need for continuous adaptation drives the need to innovate in rapid development cycles—an option traditionally not offered in the field of data science, where agility gives way to lengthy and expensive R&D projects with high levels of uncertainty. It shouldn’t come as a surprise that 87% of data science projects never make it into production.
How is Zorroa enabling organizations to adapt?
In 2020, we’ve begun to see the emergence of “Dual Track Transformation,” defined as the organization’s ability to concurrently execute long-term transitions and real-time process innovations. As media technologists who set out to disrupt the way AI and machine learning is being integrated into today’s software, the 2020 shift drove our Zorroa team to spend the year building a solution that breaks down barriers to AI/ML adoption.
Enter no-code machine learning.
Zorroa’s no-code ML-integration platform does three things to modernize the digital media supply chain.
- Kick off machine learning projects in under an hour: Customers are able to stand up machine learning projects and see results in a matter of hours, accelerating time to their first POC.
- Support interoperability: In lieu of siloed workflows and disparate tech stacks, Zorroa customers get direct access to the ML API ecosystem that includes Google, AWS, and Azure, and a better way to integrate ML-generated metadata into their content management and production tools..
- Scale rapid-cycle innovations with no-code ML: ML projects, even when using off-the-shelf ML APIs, are complex and unpredictable, requiring 10-12 months of development time and half a dozen engineers. Every click in Zorroa’s UI eliminates months of development and arms engineering teams with the tools they need to build ML-powered applications without the ML domain expertise.
So what exactly did we build?
In 2020, we adapted our ways to meet the market where it is, delivering the type of ML solutions that enable quick experimentation with a suite of ML APIs like label and face detection, OCR, speech detection, or content moderation. Instead of the start-from-the-ground, bespoke solutions that fall under the bold bets category, Zorroa unlocks the potential of machine learning APIs by enabling customers to use a combination of APIs from any supported vendor in order to run agile ML experiments and scale their innovation projects.
Together in 2020, we have released:
- Launch of our first fully GUI-driven, SaaS platform
- Support for Google Cloud Platform, AWS, and Azure
- Addition of a dozen new ML modules
- Support for custom model training
- Data visualization UI
- Faceted metadata search
- Job queue management and error handling
- Time-based metadata for improved discovery
- Release of Scrounger, Zorroa’s open source content search application
- Open source plugin integration for third party content management and production tools like Adobe Premiere
- External ISE Security Audit
- The new Zorroa website
And in 2020, we welcomed our first set of customers to the new SaaS platform. We are very grateful for our talented team and our amazing customers who have come on this journey with us as we continue to explore uncharted territories within AI and machine learning!