Choosing the right RPA and ML technologies
Gartner’s latest magic quadrant article is out, detailing 16 of their top choices for machine learning (ML) platforms. Today, I want to share a couple of tips since reviewing the Gartner report to help you sort out which product might be best for your organization. And, even more important, I want to lay out a path to more advanced AI technologies through implementing robotic process automation – what we view as the most practical and rewarding first step towards effective digital transformation.
Digital Transformation isn’t Easy: Which tech will accelerate our progress?
It’s a daunting task to select the right ML platform for your needs. In order to choose technology that will accelerate a successful digital transformation journey, have a clear vision for its purpose. Questions like the following help build that vision, especially when you don’t have the core ML capabilities in place yet.
- Why are you considering machine learning?
- What vision do you have for machine learning outcomes for your business?
- Where do you think data science will help you identify growth opportunities, accelerate revenue recognition, modernize your customer’s experience and streamline your approach to big data?
- Where are the capabilities going to come from to realize that vision?
- What should the toolset look like? How will it integrate with your platforms?
- Which tool provides the most pragmatic path to achieve those outcomes?
Not Ready for Machine Learning? Here’s a better place to start.
We believe that robotic process automation (RPA) is the best place to start when considering your path for more “intelligent” technologies as part of your digital transformation goals.
RPA is a viable, first digital transformation enabler. The platforms are easy to learn, require lower cost resources to scale capabilities, and integrate well with the way the hype leaders (Google, Amazon, Uber, etc.) are doing data science.
Plus, given that the core foundations of machine learning and artificial intelligence exist within RPA platforms, it’s a great place to start if you don’t yet have the depth or breadth in data science.
Want to learn more about intelligent automation, specifically tied to robotic process automation, click here to access our latest POV content.
RPA’s Foothold is Strong: Why it should be a part of your long-term digital strategy.
RPA is growing exponentially. For example, UIPath, our RPA strategic partner, just announced a $153M investment to complement their year-over-year growth of 800% for 2016-2017 and a 611% increase of customer in 2017.
Beyond market growth, RPA platforms are focusing their roadmaps on making them easier to use and easier to integrate with existing systems. This path helps everyone leverage mature ML and AI stacks, providing further ‘intelligence’.
This sets the stage for continual transformation that easier to manage, but also for enabling big data science platforms in your digital ecosystem. Its going to be essential to stay nimble and relevant in a world that will no longer tolerate inefficient, manual processes or large scale back office operations eroding profitability.
Creating the Digital Transformation Building Blocks: From RPA to machine learning and beyond.
Digital transformation is being defined and executed now…all around us. Creating breakout success long-term starts with a plan to meet revenue goals utilizing advanced “intelligent” technologies.
RPA is the best next step in that digital journey. If you’ve already embarked, it might be time to plan for more advanced techniques and integrating them with your business platforms for better, more timely decisions.
Machine Learning Platforms
The Gartner report highlights 16 data science and ML platforms — their strengths and cautions — to address organizations that want to take an end-to-end approach to building and deploying data science models.
Here’s what stood out to me:
Importance of Integration
It’s important to ensure that you select a solution that helps you deliver the right experience, interoperability, and performance. How you model, collect, connect, normalize, standardize, mine and automate intelligence from your data should be achievable through what Gartner calls an ‘analytics pipeline.’ Your solution should enable fast, interconnected neural networks between data sets.
Ability to Visualize Insights
The solution must have the ability to visualize insights by area — role, function, product line, enterprise. These features will enable quick, consumable analytics to drive you forward.
New Version of Old Frankenstein
You’re going to want a new version of your old ‘Frankenstein toolkit’ that only the guy with the red stapler knows how to use. Consumability is key. Disruptive tools like Kaggle (www.kaggle.com) can give you access to new datasets to integrate with just a few clicks.
Leverage Functionality “Out of the Box”
Before ‘creating’ anything, find out what you can leverage from the platform with it’s default features. So many people I’ve talked with are prepared to spend millions on customization only to learn that they’ve not used the tools the way they were intended or they’ve tried to build a ‘better version’ of the platform. Surprisingly, many clients find that most can be achieved with little configuration and, when they do make changes, they are tight and targeted.
Invest in a Practice to Develop Capability
If you’re going down the big data path, invest in a practice. To do that you may need to partner or uncover the right leader to start. I suggest that you avoid the ‘maverick’ data scientist, but create a program to develop capabilities that will reap returns for many years.
Whether you’re ready to pick out an ML platform or trying to decide where to invest in your digital ecosystem, SDLC Partners has a roadmap to help you prioritize advanced technologies built on our roots connected to Carnegie Mellon University’s data science capabilities. Contact us to discuss or invite us in for a complimentary white boarding session.