Let’s summit up: a round up of Peak’s key takeaways from Snowflake Summit 2022￼By Chris Billingham on July 8, 2022 - 5 Minute Read
Whether it was the anticipation of reuniting with fellow Data Heroes after two years of virtual events, or the excitement of immersing ourselves in all things ‘data’ for a week, this year’s Snowflake Summit was bigger and better than ever.
Peakers at the event numbered 12, and we’ve wrapped up our key takeaways and memorable moments from the week for those that missed it.
Integration is the biggest barrier to adopting a modern data stack
Leveraging up-to-date technology within a data stack can deliver benefits that include speed, efficiency and cost savings – providing businesses have the capacity to integrate it.
Conversations at Snowflake Summit highlighted just how difficult many are finding it. Most decision makers we spoke to talked about the challenge of requiring a diverse set of skills and resources internally to get the best from their stack. It’s a balancing act, especially when coupled with the management overhead of integrations and orchestration.
The majority of Snowflake customers adopted its technology because of its simplicity and ‘as-a-service’ model. We heard many people wish they could extend this ease of use to other parts of the stack, including data science, artificial intelligence (AI) and analytics.
Having supported customers deploying AI solutions for over seven years, we couldn’t agree more. It’s one of the reasons we’re so excited about our Snowflake partnership and integration – it gives Snowflake and Peak customers the opportunity to simplify the stack, delivering data warehousing and ready-made AI applications ‘as-a-service.’
Doing even more, directly on Snowflake
Data sets are getting bigger and we’re plugging more applications into them. The bigger the data set, the harder it is to move and the more applications get pulled towards it – that’s ‘Data Gravity’.
It’s a concept that cropped up a lot at the Snowflake Summit thanks to the release of Snowpark for Python and Python user-defined functions. These new features mean that a lot of the computational machine learning workload can be pushed down into the Snowflake platform – drawing applications to the data.
Peak users bridging to Snowflake can run modeling on their own infrastructure, removing the need to move big data sets and providing more sovereignty. In general terms, using Snowflake infrastructure to do more than just ‘traditional’ database transactions provides Peak customers with a larger range of optionality when it comes to developing and deploying commercial AI.
Empower data scientists to overcome the last mile
The vast majority of AI models fail to be productionized, and many of the attendees we spoke to – whether commercial leaders or technical teams – cited this as a major challenge. Summit attendees agreed it’s not just about creating a productionizable model; it’s about empowering data scientists to get models into production themselves.
It’s a problem we designed Peak to crack. Peak unites technical and commercial teams on one platform, facilitating collaboration which promotes data science empowerment, decreases time to value and speeds up the delivery of models that meet a business need.
It’s more than just building models, it’s a cultural change too
A highlight of the event was seeing our CCO, Zoe Hillenmeyer, on stage at the Everest Theater. Consolidating many of the topics we’d heard throughout the summit from a range of industry leaders, Zoe ran a bootcamp for businesses on how to adopt commercial AI.
Here are headlines:
- You can build models rapidly by simplifying your data stack and abstracting infrastructure and integration.
- Empowering data scientists to productionize their models is the key to getting meaningful value from AI projects.
- Productionization isn’t the end of the road. End users need to embrace and utilize the output of the models, and that requires a culture receptive to change and familiar with digital transformation.
Want to read more? You can download Zoe’s presentation slides below.
A great week with great people
It’s easy to forget how intense Snowflake Summit can be but it’s always a lot of fun too. Peak CEO, Richard Potter, shared his personal highlight, “Snowflake Summit was a chance for us to learn as a team. Fueled by the best Mexican food in Las Vegas, we rallied around the booth to talk to as many attendees as possible, cheered on Zoe at her presentation, and we even managed to squeeze in a game of Topgolf! Most excitingly, we’ve been able to introduce our Snowflake x Peak proposition and highlight the value it brings to customers.”