Challenges of AI in the Mainstream

More Involvement from Engineering and DevOps “I see engineering and DevOps teams becoming more and more involved with AI.

They’ve learned that AI is something that’s going to stick around and requires different approaches than previously understood.

As the containerization of ML models becomes the norm, data scientists are being seen as 21st-century developers, which will fuel the adoption and mastery of new tools and languages.

” Stephen Galsworthy, Head of Data Science at Quby Streamlining the Path to Production “I was an invited speaker for the Boston Databricks ML Workshop and presented a webinar on our use of the Databricks Unified Analytics Platform and Snowflake.

Honestly, a good number—if not a majority—of questions were about this part of the puzzle, the last mile.

So while I’m sure we’ll continue to refine our process and technology, I’m also hopeful there will opportunities to evangelize production readiness and process on behalf of Databricks.

” Stephen Harrison, Data Science Architect at Rue Gilt Groupe Linking the Promise of ML to Business Needs “Truly leveraging ML is not easy.

  It’s attractive to executives but requires a lot of work not just in algorithms but in storage, data processing, software development, etc.

The talent is hard to find, expensive and often asked to be ‘unicorns’ in their organization.

That core issue won’t go away but more vertical-specific solutions will come up and frameworks will seek to automate more of the process.

ML integration specialists will be critical in providing the link between ML and the business to ensure value is realized.

As technology frameworks make it easier to deploy, organizations will experience ML success, less by having the best data scientists and more by ensuring rapid testing of applications against the right business challenges.

” Bradley Kent, AVP of Program Analytics at LoyaltyOne Same Challenges, Different Year “The biggest barrier to deploying AI in production is organizational silos, talent and the lack of data and infrastructure that can support an end-to-end pipeline.

 The good news is that most businesses are digitizing their processes, collaborating with partners such as Databricks and realigning organization to facilitate AI deployment.

” Mainak Mazumdar, Chief Research Officer at Nielsen The Integration of ML Expertise Throughout the Business “At Overstock, we’re continually refining how teams work together.

To that end, we have integrated machine learning within every facet of the organization.

The importance of having an AI or ML specialist reporting to the CEO is also really important for two reasons.

First, it gives AI and ML a seat at the table, bringing ideas of how this technology can be intertwined in a variety of departments, and secondly, it shows the rest of the organization that ML and AI are crucial to innovation and in order to compete in a fast-moving industry.

” Kamelia Aryafar, Chief Algorithm Officer at Overstock Progress is being made and it looks like AI is at a stage of understanding and adoption that it is crossing the chasm and gradually becoming widely adopted across all business and industries to support a myriad of use cases and business needs.

The key difference from years past is the readiness of enterprises to put ML models and AI technology into production at a greater scale than ever before.

In the 4th installment of this blog series, we will explore this topic further, as various companies will share their progress with AI and predictions on how they think AI will impact their own business in 2019.

Try Databricks for free.

Get started today.

. More details

Leave a Reply