Many teams can build AI agents as experiments, but turning them into something that works reliably, safely, and at scale for real business use is much harder. To solve this, the Vertex AI Agent Builder from Google Cloud helps organizations build, run, monitor, and secure AI agents from start to finish. In short, Google is making it easier to go from "we built a demo agent" to "we run trusted AI agents in production."

Serverless Computing

Serverless computing, often referred to as "serverless," is a cloud computing model where developers can build and deploy applications without having to manage the underlying server infrastructure. In a traditional server-based architecture, developers need to provision, configure, and manage servers to run their applications, which can be complex and time-consuming.

In a serverless architecture, the cloud provider (such as Amazon Web Services with AWS Lambda, Microsoft Azure with Azure Functions, or Google Cloud with Google Cloud Functions) abstracts away the server management aspect. Developers can focus solely on writing code for the specific functions or tasks their application needs to perform without worrying about server provisioning, scaling, or maintenance.

Serverless services offer several significant benefits that can have a positive impact on application development, deployment, and management. Some of the key significances of serverless services include:

1. Simplified Infrastructure Management
2. Auto-Scaling
3. Rapid Development and Deployment
4. Event-Driven Architecture
5. Reduced Administrative Overhead
6. Global Scalability
7. Resource Optimization
8. Improved Fault Tolerance

Amazon Web Services (AWS) continuously introduces new capabilities and features to their serverless services for several reasons aimed at improving the developer experience, expanding use cases, and meeting evolving customer needs.

Latest buzz in AWS serverless services

1. General availability of AWS Database Migration Service Serverless

On June 2nd, 2023, AWS unveiled the widespread availability of AWS Database Migration Service (AWS DMS) Serverless. This release simplifies database migrations by automating the provisioning and scalability of migration resources. With AWS DMS Serverless, users gain the ability to seamlessly replicate data across a diverse range of widely used databases, analytics engines, and services—think PostgreSQL, MySQL, Oracle, Amazon Redshift, Amazon DynamoDB, Amazon Aurora, and more. By handling the often-cumbersome work of database migration, AWS DMS Serverless minimises the need for manual resource estimation, provisioning, monitoring, and scaling. This advancement translates to migration timeframes measured in hours, and cost savings realised through payment solely for consumed data migration resources.

2. Provisioned Concurrency for Amazon SageMaker Serverless Inference

As of May 10th, 2023, AWS has introduced the general availability of Provisioned Concurrency support for Amazon SageMaker Serverless Inference. This innovative feature ensures that models deployed on serverless endpoints offer consistent performance and impressive scalability. Through the integration of provisioned concurrency, users can infuse their serverless endpoints with a predetermined volume of concurrency, effectively maintaining the readiness and responsiveness of SageMaker endpoints. This offering particularly suits scenarios where traffic is predictable, yet throughput remains relatively low.

3. Amazon Aurora Serverless v2 is now available in 4 additional regions

The footprint of Amazon Aurora Serverless v2 has now expanded to include four additional regions. Aurora Serverless v2, an adaptive, automatic scaling configuration for Amazon Aurora, has the remarkable ability to instantaneously scale to accommodate even the most resource-intensive applications. By making precise capacity adjustments, Aurora Serverless v2 ensures that an application always receives the optimal database resources it demands. This dynamic resource allocation extends to encompass an impressive range of Amazon Aurora features, spanning read replicas, multi-AZ support, Performance Insights, and Global Database functionality. This powerful suite is ideally positioned to serve a diverse spectrum of applications. Enterprises dealing with an extensive array of applications or Software as a Service (SaaS) providers managing multi-tenant environments replete with numerous databases can harness the capabilities of Aurora Serverless v2 to deftly manage database capacity across their entire infrastructure.

4. AWS Lambda introduces response payload streaming

AWS Lambda, the bedrock of serverless computing, has ushered in an exceptional enhancement: response payload streaming. This capability enables AWS Lambda functions to gradually stream response payloads back to clients, even accommodating payloads that surpass the 6MB threshold. A monumental leap forward for web and mobile applications, this feature marks a departure from the conventional request-response model. Previously, applications built on Lambda necessitated the complete generation and buffering of responses before they could be sent to clients—an approach that often resulted in delayed first-byte transmission times. With response payload streaming, Lambda functions can transmit partial responses to clients as they are ready, substantially improving the all-important first-byte transmission time, a facet crucial for web and mobile applications alike. This innovation stands to elevate the performance of AWS Lambda-powered applications to new heights.

As AWS continues to introduce new capabilities, it's evident that the serverless paradigm is here to stay. This shift isn't just a technological trend; it's a fundamental change in how we architect, deploy, and scale applications. With each enhancement, AWS reinforces its commitment to providing customers with the tools they need to succeed. With AWS's ongoing dedication to pushing boundaries, the future of cloud computing holds the promise of even more remarkable advancements.

Beinex+ AWS Offerings

AWS provides security services such as AWS Shield and AWS WAF to help protect against phishing attacks. Strengthen your defences by integrating robust security software from the AWS Marketplace and embracing Two-Factor Authentication (2FA). Safeguard against evolving threats like homograph phishing for a safer online experience.

Beinex is an AWS consulting partner, and we empower customers to host their BI solutions, provide security services and much more on the cloud. Our cloud migration experts bring in best-in-class stability and reliability by understanding your business strategy and working closely with you to deploy AWS infrastructure as a service.

Snowflake Data Sharing

Data sharing in Snowflake equips you to share specific objects with another Snowflake account or a designated reader account. The beauty of this process lies in the fact that the data isn't duplicated or moved between accounts.
Now, why is this a game-changer for organisations? When constructing data pipelines and developing data products, it's a common practice to shuttle data between databases and diverse systems to blend different datasets.
Consider this scenario: You have transactional data within your online transactional processing (OLTP) database, and you wish to integrate it with external data for a machine learning model. Traditionally, organizations would export data into a data lake, import external data, and then employ tools like Apache Spark for analysis.
But what if, instead, you could simply deposit your data into Snowflake, and the external data source could seamlessly share its data with your organisation, eliminating the need to load it separately? This eradicates the challenge of keeping data copies synchronised, resulting in savings on storage, computing costs, and maintenance efforts.
Imagine your company possesses valuable information that can guide other companies in making informed decisions. For instance, let's say your company can provide precise estimates for product delivery times based on proprietary data, and you want to offer this information for sale to your customers.
Enter Snowflake data sharing—it empowers you to precisely do that.

Case Studies: Snowflake Data Sharing

Citing two instances where leading organisations use Snowflake to improve actionable data sharing, collaboration and reporting capabilities.

1. A Pioneering Technology Leader A well-known Swedish-Swiss multinational corporation successfully implemented a streamlined data strategy using the Snowflake Data Cloud, adopting an "extract once, use everywhere" approach that simplified data consolidation and enablement. By transitioning from nightly extracts, which caused significant system overhead, to a single, near real-time Change Data Capture (CDC) process, the company achieved efficient replication of information to Snowflake with minimal impact. The utilisation of Snowflake Secure Data Sharing facilitated secure and governed data collaboration across the four business areas.
2. A Leading fast-food Restaurant Chain Snowflake's data-sharing capabilities have revolutionised decision-making for a fast-food restaurant chain. They can effortlessly share crucial sales, inventory, and operational data with external entities, expanding from three to over 30 parties.
With a high-performance database platform hosting over 2 million transaction records, the restaurant chain has established a robust data management and analysis infrastructure through Snowflake, empowering its operational and marketing endeavours.
Moreover, by consolidating all data onto Snowflake, the organisation has achieved a remarkable 70% reduction in operational IT costs, demonstrating the platform's efficiency and cost-effectiveness.
Centralising and sharing data with Snowflake significantly eased the development of data products for various purposes, including marketing campaign analytics, quotation success metrics, production line tools, and supply chain dashboards. These data products are utilised by thousands of users globally, including internal stakeholders and external vendors, enhancing collaboration and efficiency across the organisation.

What are the best practices for Snowflake data sharing?

Optimize your Snowflake data sharing experience with these essential practices. Ensure data security by utilizing secure views to filter and mask sensitive information. Enhance clarity and understanding by employing descriptive names and comments for your shares. Monitor and fine-tune your sharing activities using Snowflake Information Schema or Account Usage views. Foster communication and collaboration with your consumers to create a seamless workflow.
Take command of your data sharing environment by setting quotas and limits with the ALTER SHARE command. Keep your consumers informed about any changes or updates to your shares, and actively seek feedback to refine your data-sharing strategy. Explore additional data sources through Snowflake Data Exchange or Data Marketplace to enrich your analytics.
These best practices safeguard sensitive data, ensure compliance with data privacy regulations, clarify the purpose of each share, and provide insights into usage and performance, ultimately enhancing your data analysis capabilities. Below are some best practices for data sharing with Snowflake:

1. Understand Snowflake Data Sharing
Familiarize yourself with Snowflake's data sharing features, such as Secure Data Sharing (SDS) and Sharehouse, to leverage the platform effectively.

2. Role-Based Access Control (RBAC)
Implement strong RBAC policies to control who can share data and who can access shared data. Define roles and permissions to ensure data security and compliance.

3. Secure Data Sharing
Use Secure Data Sharing (SDS) to securely share data with external parties without copying or moving the data. Implement encryption and access controls to protect sensitive information.

4. Sharehouse Best Practices
If using Sharehouse, follow best practices for creating and managing share objects. This includes defining share schemas, tables, and using the appropriate share options for your use case.

5. Data Masking and Redaction
Apply data masking or redaction policies to shared data to protect sensitive information. Ensure that shared data complies with privacy regulations and internal data governance policies.

6. Query Performance Optimization
Optimize query performance for shared data by using clustering keys, partitioning, and indexing. This helps enhance the efficiency of queries on large datasets.

7. Versioning and Change Tracking
Implement versioning and change tracking mechanisms to keep track of updates and changes in shared data. This ensures data lineage and helps with auditing and troubleshooting.

8. Documentation and Metadata
Maintain comprehensive documentation and metadata for shared datasets. Include information about the source, purpose, and any transformations applied. This helps users understand the shared data context.

9. Governance and Monitoring
Establish governance practices for data sharing, including regular reviews of shared data objects and access logs. Monitor data-sharing activities to identify any anomalies or potential security issues.

10. Educate Users
Provide training and documentation for users involved in data-sharing activities. Ensure they understand the best practices, security protocols, and the impact of data sharing on performance.

11. Regular Audits and Reviews
Conduct regular audits and reviews of shared data objects, permissions, and access controls. This helps maintain data integrity, security, and compliance with organizational policies.

12. Cost Monitoring

What’s Next

1. Enhanced Data Collaboration Tools:

Best Way to Share Data for Your Business

For secure collaboration, old ways of copying data are no longer the best. If you're working with trusted partners and it's privacy-compliant, Snowflake Secure Data Sharing is a quick and secure option. But, if you're dealing with sensitive or regulated data, especially when the risk is high, consider using a data clean room for an extra layer of security and compliance.

Beinex + Snowflake Offerings

Beinex’s partnership with Snowflake enables us to offer you advanced features like automated tuning and elastic compute, along with analytics modernisation services, to help your organisation realise exponential Return on Investment.

Note: The Server Management Add-on is not available for Tableau Online, as they manage everything from scaling, performance, and security on behalf of their Tableau Online customers. The Tableau Server Management Add-on can be separately purchased from the Tableau Server deployment.

To start, access the Superstore Dataset, available for download on Kaggle. Use Tableau Public to connect to this dataset and explore the rows and columns to understand the documented data related to the Superstore. This exploration sets the stage for the subsequent cohort retention analysis. Move over to the Sheet 1 tab; this will serve as your dedicated workspace where you'll craft your visualizations. This area will be your canvas to build and design the subsequent analyses and visual representations.

To perform cohort analysis effectively, specific data points are necessary:

Creating Calculated Fields

One crucial calculated field needed is the "Customers’ First Purchase Date (quarter)." Since the dataset doesn't contain the first purchase date field, a calculated field is essential. This field is derived from computations based on existing data in the dataset, enabling the identification of when each customer made their initial purchase. Calculated fields are instrumental in manipulating and analyzing data by performing calculations based on existing dataset information.

To create quarter and year cohorts based on the customers' first purchase date:

Assembling the cohort retention table involves using the calculated fields previously created to form a comprehensive table that illustrates cohort-based metrics, particularly the number of unique customers per first quarter and the retention rate.

Here's how to assemble the cohort table:

This table allows for a clear visualization of how customer retention varies across different cohorts over subsequent periods, providing insights into the effectiveness of retaining customers acquired in specific quarters.

This step-by-step guide is incredibly detailed for constructing the cohort table in Tableau. It breaks down the process systematically, ensuring proper visualization of the cohort analysis.

Interpreting the retention rates from this table involves examining both the rows and columns: