Artificial intelligence (AI) is no longer the future, it is here, and it is now. Organizations are already building solutions based on AI, and consumers are already experiencing the power of AI every day in numerous applications. In this new data economy, the true enablers are those who empower creators and organizations to build AI applications and experiences quickly. Those are the kind of startups we’re going to feature in this article. These three startups each take a different route to provide AI solutions for enterprises, but all have something unique to contribute to the conversation.
1. Dataiku: End-to-end enterprise AI management
A haiku is “a Japanese poem consisting of three short lines that do not rhyme.” In like manner, Dataiku is out to democratize enterprise data so that it is approachable and usable by everyone in the organization. AI can be confusing for everyone, but by using a powerful AI solution, it can be meaningful, useful, and even quick.
Dataiku’s enterprise AI solution handles the entire data lifecycle from data collection and cleaning to analysis and presentation. It is a server-based solution that can be run either on-premises or in the cloud on any public cloud platform. It can access data from any of these data stores and help with data wrangling and cleaning.
Dataiku enables users to build data models using open source libraries like Scikit-Learn, MLlib, and XGboost. It also allows training models using Python and R – two favorites tools in the arsenal of every data professional.
From the get-go, Dataiku ensures all data is versioned using the industry-standard GitHub. Dataiku makes data approachable and usable by allowing employees to collaborate around data. It does this by giving each data set its own metadata such as a description, the ability for users to comment on data, and even tag data. Further, it allows users to compare data models and make complex decisions quickly.
Once you’ve decided on a particular data model, you can deploy it with just a few clicks. Here, there are multiple deployment options, and you can even deploy in the cloud on Kubernetes. Dataiku has support for GPU-based instances for high power data processing.
The power of Dataiku is that it manages the entire lifecycle of machine learning data modeling. It is a solution for data scientists, operational workers, and executive decision-makers. They can all collaborate around data on the same platform, and speak the same language, so to say. Dataiku quickens machine learning processes and adds value by making every step more collaborative.
Dataiku is one of the leading enterprise AI startups today. A testament to this is their recent $100 million funding in a Series D round. With many big-name customers and a fast-growing team of 450 employees, Dataiku is a leader in the space.
2. C3: Low-code programming for enterprise AI
C3 does this by abstracting all application components, such as data storage, data processing, machine learning models, and data security. C3 claims that it can cut short a developer’s coding time by 95 percent by leveraging this abstraction. That is a bold claim and can save large organizations a lot of time and money in their quest for AI dominance.
C3 offers three product lines — C3 AI suite, C3 AI Applications, and C3 AI Ex Machina. The first is an end-to-end solution to build AI applications in the enterprise. The second is a set of SaaS AI tools that are already built and ready to use. The third is a no-code data analytics solution.
C3 features a visual data builder with drag and drop functionality that enables developers to enjoy “declarative programming.” This is when they can simply define what they’d like to have happen, and let C3 take care of the execution of the tasks. Declarative programming is the future of development and is a key part of the broader GitOps movement. Using key development terms like “low-code” and “declarative programming,” C3 shows its keen awareness of what’s happening in the broader world of DevOps.
C3 allows deploying to any major cloud provider such as AWS, Azure, Google Cloud, and IBM. In fact, C3 recently announced partnerships with all four major cloud vendors, which will expand its reach to its customers.
C3 addresses a wide range of use cases from money laundering to CRM to energy management. One of their recent projects involved taking disparate COVID-19 datasets worldwide and combining it into one unified, organized, and cohesive data set. This saves data analysts hours of data preparation time and greatly quickens the process.
With a clear story and value proposition, C3 is well-positioned to be a frontrunner in the race for enterprise AI supremacy.
Verta: DevOps for ML
Verta is more early-stage than the previous two startups. It recently raised $10 million in Series A funding and is off to the races.
Verta is based on the open-source tool ModelDB, which handles versioning, deployment, and monitoring machine learning models. As such, Verta doesn’t handle phase 1 of the AI lifecycle, such as data cleaning and data preparation. It comes in at phase 2 — the model creation phase.
Many AI projects for enterprises don’t see the light of day simply because they get lost in a complex process from creation to deployment. Verta believes that it can streamline this process and increase the chances of AI projects being successfully deployed and managed.
Verta calls its solution “MLOps,” or DevOps for ML. Verta doesn’t concern itself with data preparation or even model creation as there are powerful tools like TensorFlow and Jupyter that have already solved the model creation problem. Once a model is created, Verta takes over by building in model versioning. It then handles model deployment and autoscaling of models in production. Finally, it helps monitor and manage models that are deployed.
For organizations that already have a data management solution and have a great process for creating ML models, Verta can complete the journey to enterprise AI by bringing in robust deployment and management of models. It is an Ops-centric solution with some benefits for developers and data scientists as well.
AI and enterprises: No looking back
It is an exciting time for enterprises looking to up their AI game. There are numerous solutions available today to take an enterprise from data to AI in no time. Every organization is at a different point in their enterprise AI journey. Some are starting from scratch, and others have already invested years building data science teams and just need to fill one particular gap in their process. Whether you opt for a top-of-the-line solution like Dataiku, a purpose-built one like C3, or a newcomer like Verta, there’s something for everyone. Enterprises will no longer be slowed down by heavy legacy applications and processes — they are all set to be modernized by the power of AI. The startups mentioned here are going to make this possible.
Featured image: Pixabay