Behind the Booth: How We Delivered a Live AI Experience at Snowflake Summit

Behind the Booth: How We Delivered a Live AI Experience at Snowflake Summit

Written by

Ryan Lewis, Sr. Lead Consultant

Published

June 13, 2025

AI & Machine Learning
Sigma
Snowflake

We wanted to do more than just hand out stickers at this year’s Snowflake Summit. We wanted to create an experience—something fun, immersive, and hands-on that brought the power of AI, Snowflake, and real-time analytics to life.

So we built a two-part AI solution that combined facial recognition, mood detection, and instant dashboarding. The result? A booth where attendees could get a custom AI-generated caricature and learn how advanced tools like Snowflake, Landing AI, and Sigma come together to power intelligent experiences.

The Big Idea: Fun Meets Function

We set out to show what’s possible when modern data and AI tools are thoughtfully combined. Our booth experience worked like this:

1

Snap a Photo
Attendees had their photo taken at the booth.

2

Generate a Caricature

The image was sent to an AI service to create a playful caricature.

3

Analyze Your Mood

At the same time, the photo was sent to a custom mood detection model powered by LandingLens (via Snowflake) to classify your expression.

4

See the Results in Real Time

Your mood data was logged in Snowflake and instantly visualized on a Sigma dashboard alongside aggregated visitor insights.

5

Take Home the Fun

Every participant got a printed caricature to keep—and a great story to share on LinkedIn.

Behind the Scenes: How It All Worked

Data Collection & Processing

We started by gathering internal data—photos of the OneSix team acting out specific moods. These images were passed through AWS Rekognition to crop the headshots and stored in an S3 bucket.

But to train a meaningful model, we needed more. We augmented our dataset with hand-picked stock photos and even generated synthetic headshots using GPT-image-1 to ensure variety and balance across moods.

Model Training in LandingLens

Training our mood detection model involved:

LandingLens made it easy to go from raw images to a trained model with its low-code interface and tight Snowflake integration.

Real-Time Inference & Dashboarding

When an attendee took a photo:

The entire process happened in seconds. One moment you’re smiling at the camera, the next you’re a dot on a live mood chart.

Why We Chose These Tools

Snowflake Icon (2)
landingai-icon

Landing AI's LandingLens

sigma (2)

Sigma

The Takeaway

This wasn’t just a gimmick. It was a live demonstration of what’s possible when you combine the right technologies with a little creativity.

At OneSix, we believe data and AI should feel approachable and practical. This booth experience proved that smart, scalable, and governed AI applications don’t have to be intimidating—or boring. Thanks to everyone who stopped by to participate. We hope you had as much fun as we did. See you next year!

AI-Ready. Future-Focused.
Powered by Snowflake + OneSix.

Whether you’re just starting or scaling, OneSix helps you build intelligent data solutions with Snowflake it's core. We bring the strategy, architecture, and AI expertise to turn your data into real business outcomes.

Click Here

AI-Driven and Privacy-First: How Snowflake Powers Modern Marketing

AI-Driven and Privacy-First: How Snowflake Powers Modern Marketing

Written by

Jacob Zweig, Managing Director

Published

May 7, 2025

AI & Machine Learning
AI-Driven Marketing
Snowflake

The rules of marketing are changing fast. AI is raising the bar for personalization. Privacy regulations are reshaping how you work with customer data. And siloed data stacks? They’re quickly becoming a thing of the past. That’s why OneSix partners with Snowflake — to help marketing teams not only keep pace, but lead.

We bring together AI, machine learning, and Snowflake’s AI Data Cloud to help companies build smarter, faster, and more personalized marketing strategies rooted in trusted, connected data. Together, we give you the tools to act on insights, optimize spend, and elevate performance — all while protecting customer privacy and scaling with confidence.

5 Trends Reshaping Marketing

Snowflake’s Modern Marketing Data Stack 2025 report highlights five major forces reshaping the landscape. But to succeed, companies need more than ideas — they need a foundation built for agility, intelligence, and security. That’s where Snowflake shines — and where OneSix delivers real business value.

Trend 1

The Rise of the Data-Empowered Marketer

In 2025, marketers are no longer waiting on technical teams to access insights. Thanks to advancements in AI and natural language querying, they can now explore data directly, make decisions faster, and create more personalized experiences​.

At OneSix, we help marketing teams take full advantage of this shift with personalization at scale, deploying models like Lifetime Value (LTV), churn prediction, and Next Best Action to deliver immediate, actionable insights that drive real-time personalization.

Trend 2

Sophisticated, Data-Connected Applications

Marketing applications are evolving to connect directly to unified data platforms rather than relying on fragmented, siloed subsets. This new model boosts security, strengthens governance, and unlocks deeper insights across every customer interaction​.

OneSix designs marketing ecosystems where applications work seamlessly with unified customer data, creating the foundation for truly cohesive, omnichannel engagement.

Trend 3

Old and New Measurement Strategies

As third-party cookies become less reliable, marketers are turning to a blend of classic and modern measurement tools, including Marketing Mix Modeling (MMM) and secure Data Clean Rooms​. These approaches offer a privacy-first way to measure performance and optimize budget allocation.

Through our Marketing ROI Measurement services, OneSix helps companies implement next-generation MMM models and privacy-centric attribution frameworks to deliver actionable clarity on campaign effectiveness.

Trend 4

The Increased Value of First-Party Data

In a privacy-first world, first-party data has become marketing’s most valuable asset. Brands that successfully capture, enrich, and activate their own customer data are gaining undeniable competitive advantages​.

At OneSix, we empower organizations to build rich, scalable first-party data ecosystems—leveraging clustering, segmentation, and predictive modeling to unlock smarter acquisition, retention, and personalization strategies.

Trend 5

The Rise of Commerce Media

More brands are transforming into media platforms themselves, monetizing their first-party data by creating targeted advertising ecosystems. Whether in retail, travel, telecom, or beyond, this shift to commerce media opens new revenue streams and deepens customer engagement​.

OneSix supports brands in navigating this opportunity with our high-value audience acquisition solutions, helping companies not only identify and convert high-value customers but also build monetizable audience strategies through predictive insights and lookalike modeling.

Snowflake Icon (2)

Snowflake: The Core of Modern Marketing

To fully take advantage of these emerging marketing trends, companies need more than ambition — they need the right foundation. As Snowflake highlights, building a future-ready marketing stack means embracing platforms that are connected, composable, and AI-powered​.

At the heart of this transformation are key capabilities and technologies that define the modern marketing data stack:

Unified, AI-Ready Data Platform

Snowflake's AI Data Cloud eliminates data silos by centralizing customer, campaign, and sales data in a single, governed environment. This "single source of truth" unlocks faster personalization, smarter segmentation, and more efficient optimizations.

Advanced AI and ML Services

With Snowflake Cortex, marketers can easily tap into pre-built machine learning models and generative AI capabilities for tasks like customer segmentation, predictive scoring, and automated personalization — without needing deep technical expertise.

Privacy-First Collaboration

Snowflake's Data Clean Rooms enable secure data collaboration with partners and media platforms, allowing companies to measure campaign performance and enrich audience insights while fully preserving user privacy.

Third-Party Enrichment

Through the Snowflake Marketplace, marketers can access hundreds of third-party data sources — from demographic and intent data to purchase behavior — enriching their own first-party data without complex integrations.

Identity Resolution and Enrichment

Marketers can tie together fragmented user profiles and anonymous interactions using Snowflake-native identity resolution tools, making it easier to create a truly holistic view of the customer journey.

Governance and Security Built In

Snowflake ensures data security, governance, and compliance at every layer, helping marketing teams maintain customer trust while deploying increasingly sophisticated personalization strategies.

At OneSix, we help companies not just implement these Snowflake capabilities — but also design strategies around them. We build AI-driven marketing ecosystems that are fueled by unified data, automated by intelligent models, and powered by real-time insights — so you can deliver the right message, to the right customer, at exactly the right time.

Whether you’re ready to deploy Snowflake Cortex for predictive engagement, leverage Data Clean Rooms for collaborative attribution, or enrich your segmentation strategies through Snowflake Marketplace, OneSix can accelerate your path to marketing success.

Future-Proof Your
Marketing Strategy

At OneSix, we don’t just implement Snowflake — we create custom, AI-powered marketing strategies built on it. From real-time personalization to privacy-first measurement, we’re here to help you lead with data, act with intelligence, and scale with confidence. Let’s reimagine your marketing strategy together.

Contact Us

Using AI to Extract Insights from Data: A Conversation with Snowflake

Using AI to Extract Insights from Data: A Conversation with Snowflake

Published

February 6, 2025

During Snowflake’s World Tour stop in Chicago, Data Cloud Now anchor Ryan Green sat down with leaders from OneSix. During the conversation, Co-founder and Managing Director Mike Galvin and Senior Manager Ryan Lewis note how Snowflake’s technology has changed the game, allowing them and its customers to focus less on how to build data infrastructure and more on how to extract insights from data, be it via the use of AI or reporting or dashboarding.

Get More from Your Data with Snowflake

As a Premier Snowflake Services Partner, we drive practical business outcomes by harnessing the power of Snowflake AI Data Cloud. Whether you’re starting with Snowflake, migrating from a legacy platform, or looking to leverage AI and ML capabilities, we’re ready to support your journey.

Contact Us

Making AI More Human: The Power of Agentic Systems

Making AI More Human: The Power of Agentic Systems

Written by

Jack Teitel, Sr. AI/ML Scientist

Published

December 13, 2024

AI & Machine Learning
AI Agents & Chatbots
Snowflake

As AI advances, large language models (LLMs) like GPT-4 have amazed us with their ability to generate human-like responses. But what happens when a task requires more than just straightforward answers? For complex, multi-step workflows, agentic systems represent a promising frontier, offering LLMs the ability to mimic human problem-solving processes more effectively. Let’s explore what agentic systems are, how they work, and why they matter.

What are Agentic Systems?

Agentic systems go beyond traditional one-shot prompting — where you input a single prompt and receive a single response — by introducing structured, multi-step workflows. These systems break down tasks into smaller components, use external tools, and even reflect on their outputs to iteratively improve performance. The goal? Higher-quality responses that can tackle complex tasks more effectively.

Why Traditional LLMs Fall Short

In a basic one-shot prompt scenario, an LLM generates a response token by token, from start to finish. This works well for simple tasks but struggles with:

For example, if you ask a standard LLM to write an essay or debug a piece of code, it might produce a flawed output without recognizing or correcting its mistakes.

One method of correcting these limitations is to use multi-shot prompting, where the user interacts with the LLM, sending multiple prompts. By having a conversation with the LLM, a user can point out mistakes and prompt the LLM to provide better and more refined output. However, this still requires the user to analyze the output, suggest corrections, and interact with the LLM more than just the original prompt, which can be rather time-consuming.

One-Shot Prompting

Multi-Shot Prompting

Categories of Agentic Systems

Agentic systems address these limitations by employing four key strategies:

1. Reflection

Reflection enables an LLM to critique its own output and iteratively improve it. For instance, after generating code, a reflection step allows the model to check for bugs and propose fixes automatically.

Example Workflow:

2. Tool Use

Tool use allows LLMs to call external APIs or perform actions beyond simple token generation (the only action within scope of a traditional LLM). This is essential for tasks requiring access to real-time information via web search or needing to perform specialized functions, such as running unit tests or querying up-to-date pricing.

Example Workflow:

3. Planning

Planning helps LLMs tackle complex tasks by breaking them into smaller, manageable steps before execution. This mirrors how humans approach large problems, such as developing an outline before writing an essay.

Example Workflow:

4. Multi-Agent Systems

Multi-agent systems distribute tasks among specialized agents, each with a defined role (e.g., planner, coder, reviewer). These specialized agents are often different instances of an LLM with varying system prompts to guide their behavior. You can also utilize specialized agents that have been specifically trained to perform different tasks. This approach mirrors teamwork in human organizations and allows each agent to focus on its strengths.

Example Workflow:

Why Agentic Systems Matter

Agentic systems offer several advantages:

Practical Applications of Agentic Systems

Coding Assistance​

In software development, agentic systems can write code, test it, and debug autonomously. For example:

Business and Healthcare

In domains where decision-making requires transparency and reliability, agentic systems excel. By providing clear reasoning and detailed workflows, they can:

Real Time Information Analysis

Many businesses, such as finance, stock trading/analysis, e-commerce and retail, social media and marketing, rely on real-time information as a vital component of their decision-making. For these applications, agentic systems are necessary to extend the knowledgebase of stock LLMs beyond their original training data

Creative Collaboration

From generating marketing campaigns to designing product prototypes, multi-agent systems can simulate entire teams, each agent offering specialized input, such as technical accuracy, customer focus, or business strategy.

Implementing Agentic Systems

Building agentic workflows may sound complex, but tools like LangGraph simplify the process. LangGraph, developed by the creators of LangChain, allows you to define modular agent workflows visually, making it easier to manage interactions between agents. Any code or LLM can act as a node (or agent) in LangGraph.

For example, if working in Snowflake, LangGraph can be combined with Snowflake Cortex to create an agentic workflow leveraging native Snowflake LLMs, RAG systems, and SQL generation, allowing you to build complex agentic workflows in the same ecosystem as more traditional data analytics and management systems while ensuring strict data privacy and security.

For simpler use cases, platforms like LlamaIndex also support agentic capabilities, particularly when integrating data-focused workflows.

The Future of Agentic Systems

As research evolves, agentic systems are expected to remain relevant, even as base LLMs improve. The flexibility of agentic workflows ensures they can be tailored to specific domains, making them a valuable tool for automating complex, real-world tasks. In addition, as base LLMs improve, you can keep your same agentic workflows in place, but swap out the individual agents for the improved LLMs, allowing you to easily improve the overall system performance. In this way, agentic systems not only improve accuracy of traditional LLMs, but can easily scale/adapt to the current rapidly changing LLM ecosystem.

In the words of AI pioneer Andrew Ng, agentic systems represent “the next big thing” in AI. They offer a glimpse into a future where AI doesn’t just respond — it reasons, plans, and iterates like a true digital assistant.

Get Started

Ready to harness the power of Agentic AI? We’ll help you get started with tailored solutions that deliver real results. Contact us today to accelerate your AI journey.

Contact Us

Data and AI Horizon 2025: Key Trends and Tips

Data and AI Horizon 2025: Key Trends and Tips

Written by

Jacob Zweig, Managing Director

Published

December 6, 2024

Data & AI Strategy
Retail & Consumer Goods
Manufacturing
Healthcare & Life Sciences
Financial Services
Snowflake

As we enter 2025, organizations face an unprecedented convergence of technological advancements in AI, computing, and human-centered innovation. This year marks a pivotal shift from experimentation to operationalization, with a focus on measurable ROI, ethical governance, and sustainable practices. Industries from manufacturing to healthcare are leveraging these trends to drive efficiency, collaboration, and customer-centric solutions.

By embracing cutting-edge tools such as autonomous AI systems, hybrid computing architectures, and AI-driven personalization, businesses can transform operations, unlock new opportunities, and thrive in a rapidly evolving digital landscape.

Drawing on insights from Snowflake’s AI + Data Predictions 2025, Coalesce’s Top Data Trends for 2025, PwC’s 2024 Cloud and AI Business Survey, and Gartner’s Top 10 Strategic Technology Trends for 2025, this guide explores key trends, industry-specific impacts, and strategic recommendations to help leaders navigate and harness the transformative potential of 2025.

Table of Contents

Top 3 Trends

TREND 1

Practical, Value-Focused AI

AI remains a cornerstone of innovation, but 2025 marks a decisive shift from exploratory projects to operationalized solutions that deliver measurable ROI.

Aligning AI with Business Goals

Over the past two years, businesses faced immense pressure to rapidly adopt AI technologies, driven by demands from investors, boards, and executives. This rush often resulted in disjointed experiments with tools like ChatGPT, revealing both the potential and the challenges of unstructured adoption. Now, companies are recalibrating their focus, aligning AI initiatives with broader data strategies for strategic, well-defined outcomes.

“People are coming to the realization that building an AI solution is very easy, but building an AI solution that actually adds value is much more difficult.”

Governance as a Foundation

Ethical AI governance is no longer optional. Transparent guardrails and accountability are critical to mitigate risks like bias and data poisoning while fostering stakeholder trust.

“You can have a use case with AI, but if you have not put the right guardrails around that and understood governance and responsible AI, then obviously you leave yourself exposed as an organization. It’s really all about governance and transparency.”

Generative AI and Automation

Generative AI and autonomous agents are transforming productivity by automating workflows, streamlining repetitive tasks, and introducing novel use cases. Tools like Retrieval-Augmented Generation (RAG) enhance reliability by grounding outputs in verifiable data, addressing the challenge of hallucinations.

67%

of top-performing companies are already realizing value in using GenAI for products and services innovation.

TREND 2

Seamless Data Architectures

Effective AI relies on data architectures that are robust, scalable, and interoperable. These architectures ensure seamless data integration and processing, enabling AI to deliver reliable and impactful outcomes.

Unified Storage for Seamless Processing

Organizations are adopting unified storage solutions that integrate with multiple compute engines, enabling consistent, efficient data processing across diverse systems.

“AI models require large amounts of clean, high-quality data to function effectively and produce accurate results. Enterprises will increasingly leverage user-friendly data integration tools to centralize data from various operational data stores to create a corpus for AI training.”

The Rise of Open Table Formats

Open-source table formats like Apache Iceberg are the future of data architecture because they provide for enhanced governance and interoperability across various data platforms. Data platform leaders like Snowflake are rapidly adding features to leverage the power of Iceberg.

“Iceberg will go mainstream and finally combine operational and analytical data.”

TREND 3

Human-Centered AI Innovation

In 2025, technology will go beyond operational efficiency to reshape how humans work, collaborate, and engage with technology. Human-centered innovation empowers individuals through intuitive systems, driving unprecedented productivity and creativity.

Intelligent Workforce Automation

In 2025, technology will go beyond operational efficiency to reshape how humans work, collaborate, and engage with technology. Human-centered innovation empowers individuals through intuitive systems, driving unprecedented productivity and creativity.

“If you talk to developers about the software development lifecycle, across the design, development and testing phases, you’ll learn that pretty much no one likes QA. Good QA is very cumbersome and time consuming. If we can offload 40% or more of the testing process to an AI-powered assistant — with human supervision and assurance — we move faster, and developers spend more time doing what they love to do.”

Enhanced Team Collaboration

AI and data platforms are fostering a new wave of collaboration. Fusion teams, which combine technical and domain expertise, are driving efficient AI applications and bridging departmental gaps. Real-time data sharing enables informed decision-making and cultivates a culture of innovation.

Personalized Experiences at Scale

AI is tailoring experiences to individual needs, from customized training programs to hyper-personalized customer engagement. These advancements elevate user satisfaction, accelerate skill acquisition, and create impactful business outcomes.

“AI will transform how brands personalize and automate every step of the customer journey. Marketers will move past manual A/B testing and static targeting, embracing ML-driven experiences that continuously learn and adapt for each user.

Industry Impacts

Manufacturing

Manufacturing will experience significant advancements with the adoption of large vision models—AI systems capable of interpreting visual inputs. These technologies will:

Financial Services

Financial services will continue to lead in AI adoption but with a focus on balancing innovation and fiscal responsibility:

Healthcare and Life Sciences

Healthcare and life sciences will adopt AI cautiously, focusing on measured applications to ensure safety and compliance:

Retail and Consumer Goods

The retail industry will focus on incremental successes with AI to address challenges and enhance customer experiences:

Strategic Recommendations

Focus on Practical Applications

Learn from early adopters and prioritize use cases with measurable outcomes.

Invest in Governance

Implement frameworks that ensure ethical AI usage and compliance with regional regulations.

Embrace Open Source

Adopt open standards like Iceberg to enhance collaboration, interoperability, and vendor independence.

Upskill the Workforce

Equip teams with the skills to leverage AI and advanced computing for strategic problem-solving.

Adapt Business Models

Align organizational strategies with emerging technologies to stay competitive.

From Insight to Impact

As we navigate the horizon of 2025, the convergence of data, AI, and innovation presents organizations with immense opportunities to transform their operations and unlock new avenues of growth. By embracing these shifts, leaders can position their organizations to thrive in a rapidly evolving digital landscape. Connect with our experts today to start building a future-ready data and AI strategy that combines innovation and practicality.

Contact Us

Iceberg Tables: The Future of Data Architecture

Iceberg Tables: The Future of Data Architecture

Written by

Ryan Lewis, Sr. Manager

Published

October 30, 2024

Data & App Engineering
Snowflake
TL;DR

Open table formats like Iceberg are the future of data architecture because they provide for enhanced governance and interoperability across various data platforms. Data platform leaders like Snowflake are rapidly adding features to leverage the power of Iceberg.

In data engineering, we ETL: Extract, Transform, Load (we still call it that, although nowadays we mostly ELT). Now ask, which of the three is most valuable? Business value is realized in the transformation process. Pulling data out of and pushing it into databases is a necessary evil that engineers spend countless hours performing. The beauty of Iceberg tables is to be able to do all of your extraction, loading, and even some transformations in one platform-agnostic location, and then access that data using any tool that you want.

What Are Iceberg Tables?

Apache Iceberg is an open table format that enables organizations to manage and query large-scale datasets stored in distributed file systems like Amazon S3, Azure Data Lake Storage (ADLS), and Google Cloud Storage (GCS) while retaining the structure and flexibility of traditional databases. By supporting SQL-like queries, schema evolution, and ACID compliance, Iceberg tables allow data teams to work efficiently at scale. Iceberg isn’t the only open table format, but it does seem to be gaining the greatest adoption and momentum of the various options.

For platforms like Snowflake, Iceberg enables seamless querying of external data without the need to import it into Snowflake. This approach provides organizations with greater flexibility and control over their data, while also optimizing cost and performance.

Why Should You Use Iceberg Tables?

Apache Iceberg prioritizes flexibility by enabling data stored in object stores to be accessed by multiple compute engines like Snowflake, giving companies freedom to choose the best tool for each task. Its nature as an open format prevents the looming risk of vendor lock-in, allowing seamless platform transitions while preserving data integrity. For years, data architects have rightfully championed the need for single sources of truth, and Iceberg tables offer strengthened governance by managing data models outside of the various data platforms, creating a truly open, agnostic, and centralized data repository.

When Should You Avoid Iceberg Tables?

While Iceberg presents its advantages, it’s not a universal solution. Compatibility, for one, remains a hurdle: the current ecosystem of data platforms and BI tools hasn’t universally embraced Iceberg, making it essential for organizations to confirm that their analytics stack can support it seamlessly. Complexity is another factor. For smaller teams or those with straightforward data demands, the sophistication of Iceberg can introduce unnecessary layers of management. In these cases, a traditional databases might better serve their needs, offering efficiency without the overhead.

Who Benefits Most from Iceberg Tables?

Large enterprises with complex data ecosystems and multiple analysis tools are prime candidates for Iceberg. The ability to manage and query data across different platforms without duplicating it provides substantial value, especially during mergers and acquisitions. In these scenarios, consolidating data platforms from multiple companies can be challenging and costly. Iceberg simplifies this process by creating a unified data architecture, allowing organizations to integrate diverse data sources seamlessly without extensive data migration or replication.

Smaller companies or data teams can also benefit from Iceberg, depending on their requirements. If they anticipate growth or need flexibility in their analytics capabilities, adopting Iceberg early on could be advantageous, as it allows them to scale efficiently and avoid vendor lock-in as their needs evolve.

Why the Shift Towards Iceberg?

The data architecture landscape is shifting as enterprises realize the costs associated with vendor lock-in and data duplication. Iceberg offers a solution that supports a unified data architecture—where data remains in a single location while being accessible across various platforms. This minimizes governance challenges and provides a single source of truth.

Major Data Platforms and Iceberg Adoption

Snowflake

Snowflake has been expanding its support for Iceberg, particularly for managing external tables. This development allows organizations to query external data directly, offering a flexible and cost-effective option for integrating with other cloud platforms.

AWS

Iceberg on Amazon S3 integrates with AWS Glue Data Catalogs, supporting multiple OTFs, including Hudi and Delta Lake, making it a versatile choice for data management in AWS ecosystems.

Google Cloud

Recently, Google announced BigQuery tables for Iceberg, further evidence that Iceberg is becoming central to modern data architectures.

Microsoft Azure

Currently, Azure’s ecosystem remains more focused on Delta Lake. However, the growing demand for Iceberg may prompt future support developments.

Key Considerations for Implementing Iceberg Tables

Performance

Querying native tables has better query performance than Iceberg tables. If performance is a major consideration, you may consider creating pipelines to move some of your external Iceberg tables into your data platform. This negates the benefits of using Iceberg, so it’s important to understand your performance needs.

Cost Efficiency

Storage is inexpensive, and with Iceberg, you pay only for data storage and the compute engine used for querying and transformation. Significant savings are likely if you’re currently building redundant pipelines across platforms or frequently exporting and importing data between multiple tools, beyond the typical data extraction.

AI and Machine Learning Initiatives

Iceberg’s ability to create a consistent and flexible data source across platforms simplifies the development of AI pipelines. It reduces the need for complex ETL pipelines and minimizes data duplication, accelerating AI and ML initiatives.

Is Iceberg Right for Your Organization?​

Iceberg tables are shaping the future of data architecture, providing the flexibility, compatibility, and performance that modern enterprises need. As more platforms support Iceberg, it will likely become the standard for data lakes and warehouses, allowing organizations to break free from vendor lock-in and maintain a unified data strategy.

Get started with Iceberg tables

Book an consultation or chat with us in-person at Snowflake World Tour Chicago on November 4th.

Contact Us

Snowflake Cortex: Bringing ML and AI Solutions to Your Data

Snowflake Cortex: Bringing ML and AI Solutions to Your Data

Written by

Ross Knutson, Manager

Published

May 28, 2024

Data & App Engineering
AI & Machine Learning
Snowflake

Snowflake functionality can be overwhelming. And when you factor in technology partners, marketplace apps, and APIs, the possibilities become seemingly endless. As an experienced Snowflake partner, we understand that customers need help sifting through the possibilities to identify the functionality that will bring them the most value.

Designed to help you digest all that’s possible, our Snowflake Panorama series shines a light on core areas that will ultimately give you a big picture understanding of how Snowflake can help you access and enrich valuable data across the enterprise for innovation and competitive advantage.

What is Snowflake Cortex?

The Snowflake data platform is steadily releasing more and more functionality under its Cortex service. But, what exactly is Cortex?

Cortex isn’t a specific AI feature, but rather an umbrella term for a wide variety of different AI-centric functionality within Snowflake’s data platform. The number of available services under Cortex is growing, and many of its core features are still under private preview and not generally available. 

This blog seeks to break down the full picture of what Cortex can do. It’s focused heavily on what is available today, but also speaks to what’s coming down the road. Without a doubt, we will get a lot more new details on Cortex at Snowflake Data Cloud Summit on June 3-6. By the way, if you’ll be there, let’s meet up to chat all things data and AI.

ML Functions

Before Cortex became Cortex, Snowflake quietly released so-called “ML Powered Functions” which are now rebranded as just Cortex ML Functions. These functions offer an out-of-the-box approach for training and utilizing common machine learning algorithms on your data in the Snowflake Data Cloud.

These ML functions primarily use gradient boosting machines (GBM) as their model training technique, and allow users to simply feed the appropriate parameters into the function to initiate training. After the model is trained, it can be called for inference independently or configured to store results directly into a SQL table.

As of May 2024, there are 4 available ML Functions:

Forecasting

Use this ML function to make predictions about time-series data like revenue, risk management, resource utilization, or demand forecasting.

Anomaly Detection

This function looks to automatically detect outlier data points in a time-series dataset for use-cases like fraud detection, network security monitoring, or quality control.

Contribution Explorer

The Contribution Explorer function aims to rank data points on their impact to a particular output and is best used for use-cases like marketing effectiveness, program effectiveness, or financial performance.

Classification

To train a model that identifies some categorical value, like a customer segmentation, medical diagnosis detection, or a sentiment analysis.

In general, users should remember that these Cortex ML Functions are truly out-of-the-box. In a production state, ML use-cases may require a more custom model architecture. The Snowpark API, and eventually Container Services, allows users to import model files directly to the Snowflake data cloud, when they outgrow the limitations of the Cortex ML functions.

Overall, Cortex’s ML Functions provide a fast way for users to explore and test commonly used machine learning algorithms on their own data, securely within Snowflake.

LLM Functions / Arctic

Earlier this year, Snowflake made their Cortex LLM Functions generally available to select regions. These functions allow users to leverage LLM’s directly within a Snowflake SQL query. In addition, Snowflake also released ‘Arctic’ their open-source language model that is geared towards SQL code generation.

Below, direct from Snowflake documentation, shows how simple it is to call a language model directly within a SELECT statement with Cortex:

				
					SELECT SNOWFLAKE.CORTEX.COMPLETE('snowflake-arctic', 'What are large language models?');

				
			

In the first parameter, we defined the language model we want to use (e.g. ‘snowflake-arctic’), and in the second parameter, we feed our prompt. This basic methodology opens up a ton of possibilities for layering in the power of AI to your data pipelines, reporting/analytics, and ad-hoc research projects. For example, a data engineer could add an LLM function to standardize an free-text field during ETL. An BI developer could automatically synthesize text data from different Snowflake tables into a holistic 2-sentence summary for a weekly report. An analyst could build a lightweight RAG chatbot on Snowflake Streamlit to interrogate a large collection of PDFs.

Arctic

Arctic is Snowflake’s recently released open source LLM. It’s built to perform well in so-called ‘enterprise tasks’ like SQL coding and following instructions. It’s likely that Snowflake wants to position Arctic as the de facto base model for custom LLM business use-cases, particularly those that require fine-tuning.

Even more likely, the Arctic family of models will continue to grow. Document AI, which will give users a UI to extract data from unstructured data files, like a scanned PDF, directly into a structured SQL table. This feature is built on top of the language model ‘Arctic-TILT’.

Other Cortex / Future State

Naturally, Snowflake has joined the world is offering the Snowflake copilot to assist developers while they work with Snowflake through it’s web UI. Universal Search promises to offer an ‘augmeneted analytics’ experience where users can run a query by describing the intended result in natural language. While these features are exciting on their own, they aren’t a major focus for this blog.

Snowflake Streamlit provides a easy way to quickly build simple data applications, integrated with the Snowflake platform. Container Services opens up the possibility of hybrid architectures that leverage Cortex within external business application architectures. The VECTOR data type puts vector embeddings in columns alongside your structured data warehouse data, allowing for techniques like RAG that don’t require a new vector database like Pinecone.

Snowflake Cortex is far from fully materializing as a product, but seeing the foundational building blocks today helps paint a picture of a future data platform that enables companies to quickly and safely build AI tools at scale.

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.

Snowflake Unistore: Uniting Transactional and Analytical Data

Snowflake Unistore: Uniting Transactional and Analytical Data

Written by

Paul Narup, Senior Consultant

Published

May 20, 2024

Data & App Engineering
Snowflake

Snowflake functionality can be overwhelming. And when you factor in technology partners, marketplace apps, and APIs, the possibilities become seemingly endless. As an experienced Snowflake partner, we understand that customers need help sifting through the possibilities to identify the functionality that will bring them the most value.

Designed to help you digest all that’s possible, our Snowflake Panorama series shines a light on core areas that will ultimately give you a big picture understanding of how Snowflake can help you access and enrich valuable data across the enterprise for innovation and competitive advantage.

In this blog, we dive into Snowflake Unistore, a workload that delivers a modern approach to working with transactional and analytical data together in a single platform. Let’s explore the use cases and benefits.

OLTP vs. OLAP

Traditional relational database technology has historically been siloed into transactional (OLTP) or analytical (OLAP) systems.

OLTP, or Online Transaction Processing, is focused on managing day-to-day data operations as fast as possible and allowing operations from concurrent users. OLTP is often used for CRUD (create, read, update, delete) applications or other apps requiring transactional data. For example, banking systems would use OLTP for transfers or payments requiring real-time processing.

OLAP, or Online Analytical Processing, is designed to handle the complex queries and analysis required for reporting and business intelligence needs. These datasets are often much larger than transactional datasets as they contain much more historical data. Data is often stored in a star or snowflake schema to increase query performance.

How Unistore Works

Hybrid Tables allow for faster single-row operations like traditional transactional databases (OLTP) using included indexes and Snowflake enforcing primary keys. Primary keys, foreign keys, and unique constraints are all enforced on hybrid tables. These are not enforced on standard snowflake tables.

Unistore using hybrid tables can make sense in various business scenarios:

When systems require random reads (transactional) vs. reading a large range of data (analytical)
When systems require random writes (transactional) vs. large sequential writes (analytical – bulk loading)
Fetching all records for a customer (transactional) vs. receiving aggregations for that customer (analytical)

The Business Value of Unistore

By using Snowflake’s Unistore workload, your organization’s security and compliance needs can be simplified as only one system is needed. You’ll also be able to eliminate costs for the transactional system (hosting, maintaining, and backing up).

Additionally, most companies have an ETL process to move data from their transactional system to an analytical system because reporting on a transactional system can be slow and bog down the application it is supporting. Since Unistore allows for transactional and analytical to be stored in the same system, this ETL process can be simplified or removed—reducing complexity, latency, and maintenance issues for data teams.

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.

Administration: Taking Snowflake’s Git Integration and Schema for a Spin

Administration: Taking Snowflake’s Git Integration and Schema for a Spin

Written by

Dan Luszcz, Senior Manager

Published

May 14, 2024

Data & App Engineering
Snowflake

Snowflake functionality can be overwhelming. And when you factor in technology partners, marketplace apps, and APIs, the possibilities become seemingly endless. As an experienced Snowflake partner, we understand that customers need help sifting through the possibilities to identify the functionality that will bring them the most value. 

Designed to help you digest all that’s possible, our Snowflake Panorama series shines a light on core areas that will ultimately give you a big picture understanding of how Snowflake can help you access and enrich valuable data across the enterprise for innovation and competitive advantage. 

After Build 2023, we discussed some exciting announcements from Snowflake around Git integration and declarative schema management. With Git integration now in public preview, we decided to take these new features for a spin to see what the hands-on experience is like. 

The new features we are talking about today will enable Snowflake admins to integrate source control and use declarative schema to maintain their Snowflake environment using three key functions: 

Git Integration, which essentially lets you to setup a Snowflake stage that’s connected to your Git repository
EXECUTE IMMEDIATE FROM, which lets you run scripts directly from a stage (see above)
CREATE OR ALTER, which is a new DDL command that can perform a schema compare and update the target table to match the new schema in your latest definition

Git Integration

Git integration, which entered public preview in April 2024, enables you connect to a Git repository and interact with it like a Snowflake stage. I found that Snowflake’s implementation is a bit more involved compared to traditional Git alternatives since it integrates through SQL. The bulk of the setup consists of permissions—a song and dance very familiar to anyone who has spent any time in Snowflake.

The first challenge I faced when experimenting with this feature was the requirement to create the Git credentials and Integration in a Snowflake database. While unsurprising, in our case there was not any appropriate candidate database in our environment. The Git integration was going to be used to deploy our Snowflake OS Streamlit app, which is an Application Package and does not use an existing database. Ultimately, rather than use an existing database that had nothing to do with my use case, I decided to create a database and schema specifically for the purposes of the Git integration. Creating a database just for a Git repository rubbed me the wrong way, but it was far preferable to the alternative. 

Once the roles, database, schema, secret, and Git Repository were created successfully, I was able to fetch the master branch into Snowflake and list the files that had been pulled into the stage. Again, this is all done in SQL so you can leave your extensive knowledge of Git commands at the door. Developers will not be performing merges in Snowsight; this is simply for pulling files from Git into Snowflake for deployment. Once I had successfully fetched the files from Git into Snowflake, it was relatively easy to copy those files into an Application Package Stage.

In summary, the Git integration meets expectations. It’s not particularly user friendly, but that’s par for the course: Snowflake built an excellent cloud data warehouse and then decided to start adding features on top. That approach comes with pros and cons. For users who spend a lot of time in Snowflake, the mechanics of the new Git integration will feel very natural. For someone who is used to a traditional Git experience, it will take more getting used to. 

EXECUTE IMMEDIATE FROM

Once the files are in Snowflake from Git, it’s time to use them to manage your Snowflake environment. This is where the EXECUTE IMMEDIATE FROM command comes into play. The biggest limitation right now with this command is the need to manage script context—specifically, the database, schema, warehouse, and sometimes role being used to execute the script. In our use case, we have a Snowflake Native app that can be deployed to any Snowflake environment. To avoid hardcoding environment specific values into the deployment scripts, the deployment had to be broken up into multiple steps with USE statements mixed in with the EXECUTE IMMEDIATE statements. This will be resolved once Jinja templating (now in preview!) enables the scripts to be parameterized, allowing us to pass specific values in these environments as variables. 

CREATE OR ALTER

Finally, the star of the show: CREATE OR ALTER. As a reminder, CREATE OR ALTER enables declarative schema management, where developers can evolve table schema in-place over time. Rather than forcing developers to migrate schema by writing ALTER statements, which must be executed sequentially, declarative schema management is as simple as updating the table definition and letting Snowflake determine the necessary ALTER statements to update the target table to match. Declarative schema management enables clean and intuitive source control, collaboration, and automated deployments for your database schema. Unfortunately, this feature is still in Private Preview. Like the Jinja templating, we’ll have to wait for this to be released in Public Preview before we can start playing with it. These two features will be critical to fully unlocking the power of the Git integration feature.  Until then, it’s nice to have but ultimately limited in its scope. 

Leveraging Snowflake for Modern Dev Ops

In summary, the Git Integration and EXECUTE IMMEDIATE FROM features are an important step on the journey to implementing modern Dev Ops in Snowflake. Combined with the declarative schema management and templating features still in Private Preview, these capabilities will make Snowflake an even more compelling choice for businesses, even those that may not need to leverage it for its traditional cloud data warehouse role. Viewed alongside other upcoming features like Hybrid Tables, an exciting picture is forming where Snowflake can offer cutting-edge solutions for all the data needs of modern data organizations. 

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.

Data Engineering: The Backbone of Snowflake Data Cloud

Data Engineering: The Backbone of Snowflake Data Cloud

Written by

Kwon Lee, Senior Manager

Published

May 7, 2024

Data & App Engineering
Snowflake

Snowflake functionality can be overwhelming. And when you factor in technology partners, marketplace apps, and APIs, the possibilities become seemingly endless. As an experienced Snowflake partner, we understand that customers need help sifting through the possibilities to identify the functionality that will bring them the most value. 

Designed to help you digest all that’s possible, our Snowflake Panorama series shines a light on core areas that will ultimately give you a big picture understanding of how Snowflake can help you access and enrich valuable data across the enterprise for innovation and competitive advantage. 

Foundational to Snowflake is its ability to easily collect, store, and transform your data for different analytics workloads. This is Data Engineering and often will function as the backbone of all that’s possible. We’ve highlighted a few key data engineering features to help paint the picture of what you can do with Snowflake.

Collect/Store

Snowflake Marketplace

Centralizing data from applications and systems customers use internally is table stakes for any organization. What Snowflake augments to this process is the ability for organizations to easily discover and access datasets and services relevant to their business in the Snowflake Marketplace. For example, our client in the energy industry has retired complex, legacy data pipelines by leveraging marketplace-offered databases for key pricing data. This enables them to access real-time prices at their fingertips without the cost and time commitment to building and maintaining workloads internally. 

Kafka Connector

Data from applications can often be challenging to store and analyze because of its semi-structured formats like JSON. And streaming data like this to an analytics environment used to be a heavy lift for an organization. The Confluent Kafka Connector from Snowflake makes this process seamless. A client who leverages AWS MSK for their core application (which includes events and transcripts from their core application) streams them to Snowflake in a native format that allows for parsing and analysis by the engineering and data science team. Changes to the schema in Kafka do not disrupt the ingestion into Snowflake.

Snowpipe

A clear benefit of the cloud-native world we live in today is the low cost of storage. Organizations have leveraged this to land data from internal and external systems to AWS S3, Azure ADLS, or Google Cloud Storage. However, analyzing the data while in storage is not always straightforward or easy. Snowpipes offer a real-time ingestion capability to copy data in storage to the Snowflake Data Cloud. One healthcare organization that owns a portfolio of companies with different EMRs lands exported data files into a central Azure container where Snowpipes are set up to automate the file ingestion into Snowflake for merging and analyzing the data downstream.

Iceberg Tables

There are, however, times when copying data into Snowflake’s managed storage is not the best choice for an organization’s analytics use cases. Whether it is because the data is of a much large-scale, the data needs to be in your own storage for regulatory reasons, and/or the data needs to sit in an open source format for other internal processes/applications, Snowflake also has the capability to run workloads directly on your storage via Iceberg tables, making Snowflake’s core features available even when data cannot move beyond your managed cloud storage.  

Transform

Partner Ecosystem for Data Transformations

Regardless of what data is collected and stored, data transformations remain a core activity to make raw data valuable for insights and innovation. Snowflake offers a rich partner ecosystem that provides customers a spectrum of toolsets to cleanse and augment data for meaningful interpretation. Tools like Matillion Data Productivity Cloud offers a visual, self-documenting path towards creating orchestration and transformation pipelines, whereas dbt provides an open source or managed environment for building data models with a full-code experience. 

Snowflake Native Features

If a customer chooses to stay within the Snowflake Data Cloud itself for data transformation, there are features like Stored Procedures, UDFs (user-defined functions that can be written in Python, Java, Scala, etc.) and Tasks that you can leverage to build your own orchestrations and workflows to build a dataset that provides value. 

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.