Unifying metadata governance throughout Amazon SageMaker and Collibra

This submit was co-written with Vasiliki Nikolopoulou from Collibra.

Managing metadata throughout instruments and groups is a rising problem for organizations constructing fashionable information and AI platforms. As information volumes develop and generative AI turns into extra central to enterprise technique, groups want a constant solution to outline, uncover, and govern their datasets, options, and fashions.

Collibra is a extensively adopted information intelligence platform that helps organizations centralize governance workflows, outline enterprise glossaries, and implement insurance policies throughout information property. Groups use Collibra to curate enterprise context, classify delicate information, and handle entry to data according to compliance necessities.

Amazon SageMaker Catalog, a part of the subsequent era of Amazon SageMaker, offers a unified setting the place customers can register, search, and govern AI and information property. It permits organizations to arrange datasets, skilled fashions, options, and pipelines and apply metadata akin to enterprise phrases, classifications, possession, and utilization context. Amazon SageMaker Catalog is designed to help collaboration throughout roles, together with information scientists, engineers, and enterprise stakeholders.

As organizations scale their information and AI initiatives, making certain consistency and belief in metadata turns into more and more vital. Groups want a unified solution to handle glossary phrases, asset descriptions, classifications, and entry governance throughout platforms. With out this consistency, it turns into troublesome to implement requirements, help compliance, and drive collaboration throughout groups constructing and consuming information.

To handle this problem, Amazon Net Providers (AWS) and Collibra have constructed a brand new built-in answer that demonstrates the mixing between the Collibra Platform and the subsequent era of Amazon SageMaker. Developed collaboratively by each firms, the answer is predicated on the APIs of each merchandise and is designed to assist prospects discover what’s attainable by hands-on testing. It offers a sensible instance of how metadata synchronization between Collibra and SageMaker will be achieved in real-world eventualities. With this integration, you’ll be able to align enterprise and technical metadata throughout each platforms, so you’ll be able to prolong your governance workflows to AI and analytics property managed in Amazon SageMaker.

This answer permits metadata to stay constant throughout each platforms, no matter the place it was created. It helps scale back duplication, enhance metadata high quality, and be certain that enterprise context travels with information and AI property all through their lifecycle. The combination helps metadata synchronization, glossary time period mapping, and entry approval workflows utilizing native APIs and automation.

On this submit, we take a better take a look at the mixing, describe the use instances it permits, stroll by the structure, and present methods to implement the answer in your setting.

Resolution overview

The combination between Amazon SageMaker Catalog and Collibra gives automated, bidirectional metadata synchronization and entry governance throughout each platforms. It’s constructed utilizing the built-in APIs of Amazon SageMaker and Collibra Information Governance Middle (DGC) to offer a scalable and configurable mechanism for metadata trade. The answer consists of two major capabilities: metadata synchronization and entry subscription workflow integration. The next diagram illustrates the answer structure.

Metadata synchronization

Many organizations handle enterprise and technical metadata throughout a number of programs. With out synchronization, glossary phrases, asset descriptions, and classifications can change into inconsistent, resulting in duplicated work and misalignment throughout groups.

This integration permits metadata to movement between Amazon SageMaker Catalog and Collibra, no matter the place it was created. Key components akin to glossary phrases, their hierarchy, related descriptions, and relationships to property like datasets or columns are mechanically synchronized between platforms.

The answer helps:

Bidirectional synchronization of glossary phrases and descriptions
Preservation of glossary construction, together with parent-child relationships
Affiliation of phrases with information property akin to datasets, tables, and columns
Synchronization of extra enterprise metadata, akin to classifications and information classes
Alignment of technical descriptions for datasets and columns between programs

By holding metadata constant, the mixing reduces handbook work, avoids duplication, and offers customers in each platforms with the identical trusted context.

Subscription and approval movement

Organizations that depend on Collibra for entry governance can now prolong these workflows to property cataloged in Amazon SageMaker. After metadata is synchronized, customers can uncover and request entry to datasets immediately from inside Collibra, utilizing acquainted approval processes.

This integration connects Collibra’s workflow engine with the entry management mechanism provided by Amazon SageMaker. When an asset is registered in Amazon SageMaker and shared into Collibra, customers can provoke a subscription request in Collibra. When it’s accepted, entry is granted utilizing Amazon the SageMaker built-in entry administration, which helps a number of AWS providers akin to AWS Glue and Amazon Redshift.

Key capabilities embrace:

Discovery and entry request initiation from Collibra or Amazon SageMaker
Centralized evaluate and approval processes managed inside Collibra
Entry provisioning utilizing the Amazon SageMaker grant mechanism
Constant metadata and asset context accessible all through the request lifecycle

This movement helps streamline the expertise for each enterprise and technical customers whereas holding entry to ruled information traceable, auditable, and aligned with organizational insurance policies.

Conditions

To carry out the answer, you want the next conditions:

Walkthrough

The following part offers a walkthrough that exhibits how the mixing works from begin to end. It highlights how a person discovers an information asset, submits a subscription request, and the way that request is reviewed, accepted, and fulfilled. All through the method, metadata and governance insurance policies stay aligned between Collibra and Amazon SageMaker Catalog. This instance helps illustrate what the mixing permits and the way it matches right into a typical information entry workflow.

Setup on the Collibra setting

To allow this answer, some preliminary setup is required within the Collibra setting. This entails configuring the important thing elements that customers might want to uncover, request, and handle entry to information. The next steps define the fundamental setup required to help the general expertise.

Working Mannequin modifications and import workflows in Collibra

The working mannequin of the Collibra occasion wants two new asset varieties and attribute varieties in addition to two new relations and statuses for the scripts and workflows to work correctly. These new asset varieties are beneficial as a result of Amazon SageMaker introduces its personal ideas and structure, akin to domains and tasks. Utilizing the identical names in Collibra makes it simpler for customers to grasp and navigate each programs constantly. Within the following diagram, the brand new asset varieties are proven with dotted traces together with the corresponding new relations, attributes, and statuses.

Along with AWS tasks, the implementation requires synchronization of AWS customers past the usual capabilities. That is vital as a result of in AWS, a person can’t subscribe to an asset immediately as a person. They will solely achieve this as a member of a venture. In consequence, when a person subscribes to an asset, they have to specify which venture they’re subscribing by. To help this habits, membership to tasks data for AWS customers must be maintained and synchronized inside Collibra. AWS venture to person mapping must be maintained in Collibra, which is accessed by administrative customers. The metadata details about AWS person membership to tasks will be saved in a Collibra setting or group, which isn’t accessible to anybody besides approved customers. Steps for implementation of Collibra working mannequin modifications:

Go to Settings, then Working mannequin, and add two new asset varieties, AWS Undertaking and AWS Consumer.
In Settings, navigate to Attribute varieties and add the brand new attribute varieties. The brand new attribute varieties are: Undertaking id assigned to the AWS Undertaking asset kind, Membership to Undertaking assigned to the AWS Consumer, AWS Undertaking id, Consuming Undertaking and Consuming Undertaking Id to be assigned to the prevailing asset kind Information Utilization. Confer with the documentation for extra particulars on methods to add new attribute varieties and methods to assign them to asset varieties
In Settings, go to Relation varieties and add the Asset for use relation between asset varieties information utilization and information asset. Confer with the documentation for steerage on methods to add a brand new relation to a pair of asset varieties.
In Settings, go to Statuses and add the new statuses, that are Entry granted and Pending, to be assigned to the asset kind information utilization.
Return to the Working mannequin and, for every new asset kind, add the newly created relations, attributes, and statuses. Don’t skip this step. If it isn’t accomplished, the brand new configurations will gained’t take impact.
Create the next domains:
1. AWS Customers – This can be a enterprise asset area the place the metadata for AWS person memberships shall be saved. Customers and their memberships are mechanically imported into Collibra by the answer. An instance is proven within the screenshot.
2. AWS Initiatives – That is additionally a enterprise asset area the place AWS tasks and their metadata shall be mechanically imported. The next screenshot exhibits an instance of such a website. The AWS tasks, together with their revealed property, are introduced into Collibra by the answer.
3. AWS Subscription Requests – This can be a area of kind information utilization registry. It should maintain all new AWS subscription requests together with their context, such because the consuming venture and the subscribed information asset. The standing of every request is particularly vital as a result of it drives the mixing workflow that customers can use to trace the present state of their request.

Workflows set up

This answer consists of two workflows: one for managing subscription request approvals and one other for notifying customers when entry is granted.

The primary workflow handles the total subscription course of. It begins by prompting the person to pick the consuming venture as a result of solely tasks the person is a member of are eligible for subscriptions. After it’s chosen, a brand new subscription request asset is created in Collibra with a timestamp, the consuming venture particulars, and a standing set to Pending.

An approval activity is then assigned to the enterprise steward of the requested information asset. If the steward approves the request, the standing modifications to Accepted. This triggers a notification to the requester and alerts the AWS answer to choose up the request and grant entry. When entry is granted, the standing is up to date to Entry granted.

If the steward rejects the request, the standing is modified to Rejected and the requester is notified. No additional motion is taken in that case.

The second workflow notifies the requester that the entry was granted. It’s triggered by the capabilities in AWS when the subscription grant is accomplished. The steps to deploy the 2 workflows are as follows:

Go to Settings, then choose Workflows adopted by Definitions, as proven within the following screenshot.

Select Add a file, as proven within the following screenshot. Then, add each workflow information from the GitHub listing the place all of the information are offered. In that GitHub listing, there’s a listing with the workflow information known as Workflows.

After the workflows are uploaded, full the next steps for every one, as proven within the following screenshot:
1. Allow the workflow by selecting Play. When enabled, the button will show a Pause icon.
2. Below Guidelines, set it to use to Belongings, then select Add Guidelines and select Asset: Desk. It’s also possible to use Information Asset for a broader scope, however on this case, revealed property in AWS are tables.
3. Clear This workflow can solely run as soon as on the identical time on a particular useful resource. This offers that a number of customers can request subscriptions to the identical asset concurrently.

The workflows are actually uploaded, enabled, and prepared to be used.

Add tasks

We have to assign enterprise stewards to the ingested AWS property in order that when the workflows are triggered, there’s a designated person liable for approving subscription requests. On this model of the answer, it’s assumed that every asset has just one Enterprise Steward.

So as to add a Enterprise Steward, observe these steps:

Within the area or group the place the AWS information property have been ingested utilizing the Edge integration, select Tasks. Then select Add, as proven within the following screenshot

Select Enterprise Steward from the Position dropdown listing, as proven within the following screenshot. From the Customers or teams dropdown listing, select the person who shall be liable for approving subscription requests for these property. This answer permits just one enterprise steward per asset. You possibly can assign a enterprise steward on the group stage, and this fashion this function shall be inherited to all property beneath this group.

Select Add, as proven within the following screenshot. This can assign the chosen person to the Enterprise Steward function for the desired asset, area, or group of property.

Setup on the AWS setting

Now that the configuration on the Collibra aspect is full, arrange the Amazon SageMaker area that’s used for this walkthrough. We offer the next property to assist customers arrange this answer

An AWS CloudFormation template in YAML format, known as template.yaml
Directions to generate a lambda zip file that incorporates all of the scripts that the Cloud Formation will run, known as lambda_build.zip
Directions to create a secret utilizing AWS Secrets and techniques Supervisor that can retailer Collibra credentials.

Create the CloudFormation stack

To help this answer, provision a set of AWS sources that facilitate communication between environments and automate key duties. On this part, we present methods to deploy the foundational infrastructure utilizing AWS CloudFormation, which simplifies useful resource provisioning and offers consistency throughout environments.

On the AWS Administration Console, navigate to CloudFormation and select Create stack, then select With new sources (customary), as proven within the following screenshot.

Select the offered CloudFormation template and select Subsequent.

Enter a reputation for the stack and full all required parameters beneath:

CollibraAwsProjectAttributeTypeId – The attribute kind ID for AWS tasks in Collibra.
CollibraAwsProjectDomainId – The area ID for AWS tasks in Collibra.
CollibraAwsProjectToAssetRelationTypeId – The relation kind ID between AWS tasks and property in Collibra.
CollibraAwsProjectTypeId – The sort ID for AWS tasks in Collibra.
CollibraAwsUserDomainId – The area ID for AWS customers in Collibra.
CollibraAwsUserProjectAttributeTypeId – The attribute kind ID for AWS person tasks in Collibra.
CollibraAwsUserTypeId – The sort ID for AWS customers in Collibra.
CollibraConfigSecretsName – The identify of the AWS Secrets and techniques Supervisor secret containing Collibra configuration and credentials.
SMUSProducerProjectId – The venture ID in SMUS that incorporates the info property to be shared (producer aspect).
SMUSConsumerProjectId – The venture ID in SMUS the place shared information property shall be accessed (shopper aspect).
SMUSDomainId – The distinctive identifier for the SageMaker Unified Studio (SMUS) area.
CollibraSubscriptionRequestCreationWorkflowId – The distinctive identifier for the Collibra workflow that creates subscription requests in Collibra.
CollibraSubscriptionRequestApprovalWorkflowId – The distinctive identifier for the Collibra workflow that approves subscription requests in Collibra.
LambdaCodeS3Bucket – The S3 bucket containing the Lambda perform deployment package deal.
LambdaCodeS3Key – The S3 key (path and filename) of the Lambda perform deployment package deal throughout the specified bucket.

Choose the acknowledgement checkbox, then select Subsequent, as proven within the following screenshot.

Select Submit to begin the stack deployment. When the method is full, the stack standing will replace to CREATE_COMPLETE.

Configure shopper and producer tasks

For this submit, solely two tasks are used: one serving because the producer and one as the patron. Future variations of the answer are deliberate to help all tasks.

On the AWS Administration Console, go to the SMUS Area element web page. Below the Customers part, select Add, then choose Add IAM customers.

From the dropdown, choose the SMUSCollibraIntegrationAdminRole created by the CloudFormation template, then select Add person(s), as proven within the following screenshot.

Open the Unified Studio portal for this area and navigate to the Producer Undertaking. Go to the Members tab and select Add members.
Seek for SMUSCollibraIntegrationAdminRole and choose it from the outcomes.

Set the function to Proprietor, then select Add members.

Repeat the identical steps for the Shopper Undertaking. After including the member, the configuration ought to appear to be the instance within the following screenshot.

Be sure the producer venture has the mandatory authorization to create glossary phrases within the area unit it belongs to. For extra data, seek advice from Area items and authorization insurance policies in Amazon SageMaker Unified Studio within the Amazon SageMaker Unified Studio documentation.

Synchronization of metadata

Metadata synchronization between Collibra and SageMaker Catalog occurs on two distinct ranges, every serving a particular goal.The primary stage focuses on technical metadata. Collibra connects to providers akin to Amazon Redshift and AWS Glue utilizing JDBC and different supported connection strategies. By these connections, it ingests schema particulars together with tables, columns, and information varieties. This helps technical groups keep visibility into the construction of the datasets accessible in SageMaker Catalog.The second stage, which is the main target of this answer, handles enterprise metadata synchronization. Utilizing Collibra APIs, SageMaker Catalog retrieves enterprise glossary phrases, column descriptions, asset definitions, and the relationships amongst them. Moreover, Collibra ingests details about SageMaker tasks, the property revealed inside them, and venture membership particulars. This helps approval workflows and helps handle subscriptions primarily based on project-level entry. The next diagram illustrates how these two ranges of metadata synchronization work collectively to bridge technical and enterprise views throughout each platforms.

For the technical metadata ingestion from AWS to Collibra, observe these steps:

Inside the Collibra Edge website, create a brand new connection for every kind of AWS information retailer you wish to ingest metadata from. For detailed directions, seek advice from the About Edge and Collibra Cloud website connections within the Collibra Documentation.
1. Relying on the kind of connection, particularly if it’s JDBC, you would possibly want so as to add a functionality akin to JDBC catalog ingestion. Confer with the official documentation for extra particulars.
2. So the mixing works accurately, identify all of your AWS connections in Edge with “AWS” initially of the identify. The combination script depends on this naming conference to precisely establish property that originate from AWS.
In Collibra, go to Catalog, choose your connection, configure the principles on your schemas (akin to which tables to incorporate or exclude), and run the synchronization. It’s also possible to schedule the synchronization to run mechanically at intervals outlined within the person interface.
When metadata ingestion is full, go to Catalog, then Information Sources. You possibly can optionally filter by a particular AWS supply or maintain the default view to view all sources. From there, you’ll be able to evaluate the schemas, tables, and different metadata imported from AWS, as proven within the following diagram.

These information property are imported utilizing the JDBC connections which might be accessible from Collibra Edge. The AWS answer we current right here, along with these information property, will import AWS tasks and can hyperlink them to the property ingested right here which might be revealed in these tasks.

Technical and enterprise stewardship in Collibra

Collibra offers enterprise glossaries to outline enterprise context. These glossaries may also embrace a hierarchy or taxonomy of enterprise phrases primarily based on their interdependence. The next is an instance of a glossary used for this submit.

An Order consists of elements akin to Order Date, Order ID, and others. In Collibra, Enterprise and Technical Stewards are liable for linking Enterprise Phrases to the columns and tables ingested from AWS, as proven within the following diagram. For detailed steerage on methods to carry out stewardship actions, seek advice from the official Collibra documentation.

All the enterprise glossary with its one-level hierarchy is imported into AWS SageMaker Unified Studio mechanically with this answer. Some enterprise phrases are additionally linked to information classes which might be related to information privateness, regulatory insurance policies, and requirements. Within the instance within the following screenshot, buyer ID is related to a knowledge class. This connection between enterprise phrases and information classes hyperlinks the related information to related insurance policies and requirements. In consequence, a desk or column related to a enterprise time period that’s linked to a knowledge class may also inherit the related coverage or customary.

The enterprise time period buyer ID is linked to the info class personally identifiable data (PII). With this relation, all columns or tables which might be linked to this enterprise time period mechanically inherit the PII information class, and due to this fact the insurance policies linked and related to it.

The metadata is imported into AWS SageMaker Unified Studio on the asset and schema ranges.

All of the enterprise metadata described beforehand is synchronized with AWS utilizing this answer. Descriptions, information classes, tags, enterprise phrases are all imported into AWS and linked to respective property. Within the README, the info class is related to one of many columns and the enterprise time period related to a desk or dataset.From Collibra we import into AWS the next:

Enterprise phrases and their hierarchies and descriptions
The hyperlink of the enterprise phrases to the technical property
Information class of enterprise phrases inherited within the technical property imported within the README part of the technical asset
Tags and descriptions of technical information property

Not solely is the enterprise time period imported into AWS SageMaker Unified Studio, its taxonomy is imported precisely as it’s in Collibra. The next screenshot exhibits an instance the place order is imported to have beneath it the enterprise phrases order ID, amount, and so forth.

Subscription to revealed property

For the subscription course of, the identical workflows and collection of duties happen whether or not the request is initiated from AWS or from Collibra. An summary of those duties and the end-to-end movement from each platforms is proven within the following diagrams:

This diagram outlines the subscription request movement when initiated from Collibra. A person searches for a enterprise time period, locates the associated asset, and submits a subscription request. The system creates a corresponding request asset in Collibra. The person then selects the vacation spot venture for the info. An approval workflow is triggered, notifying the designated enterprise steward. If the request is accepted, SageMaker Catalog mechanically provisions entry and updates the request standing to Entry Granted. The person receives a ultimate notification confirming entry. This course of offers managed, clear information sharing throughout platforms.

The next diagram illustrates the end-to-end subscription movement when the info person initiates the method from inside SageMaker Studio. The person begins by looking for information utilizing a enterprise time period and deciding on the related asset. After selecting the suitable desk, they request entry, which triggers the creation of a subscription request asset in Collibra. The person then selects a vacation spot venture primarily based on their memberships. Collibra sends an approval request to the designated enterprise steward, who critiques and both approves or rejects it. If accepted, SageMaker Catalog mechanically provisions the subscription and notifies the requester. The subscription request standing is then up to date to Entry Granted, finishing the workflow.

For this submit, the method is described ranging from Collibra, though it capabilities the identical method if initiated from AWS. On this instance, an information shopper is looking for information associated to AWS orders utilizing the Collibra interface.

In Amazon SageMaker Unified Studio, the info shopper is a member of the Orders and Merchandise venture. At this stage, the person has no energetic subscriptions or entry to information property. The next screenshot is included for instance the state earlier than the mixing takes impact.

In Collibra, navigate to the Search space and enter a business-friendly time period describing what the person is searching for. On this instance, enter order.

Within the Information Market, filters akin to Enterprise Phrases will be utilized to slender the outcomes by asset kind, as proven within the following screenshot. This strategy helps customers give attention to related property by ranging from clear enterprise context, which is particularly helpful when coping with many equally named tables or columns.

Within the instance proven within the following screenshot, the enterprise time period Order is chosen, and the Diagram view is opened to show its full logical lineage. The diagram exhibits that the time period is linked to the aws_orders desk. Deciding on the desk within the diagram reveals its metadata particulars, which seem on the proper aspect of the web page. From there, customers can navigate on to the desk.

Within the aws_orders desk asset, entry will be requested by initiating an AWS subscription request. From the asset view, deciding on Actions reveals the listing of accessible workflows. On this instance, the Creation of a brand new subscription workflow is chosen to begin the approval course of.

The person should choose the AWS venture to make use of because the consuming venture for the subscription. An inventory of all tasks the person is a member of is exhibited to facilitate the choice. After selecting the suitable venture, select Ship to submit the request.

After it’s submitted, the workflow is triggered, and a activity is assigned to the enterprise steward of the asset for which the subscription is requested. A brand new subscription request can be created within the AWS Subscription Requests area with a standing of Pending, and it’s mechanically linked to the requested asset.

The brand new subscription request can be mirrored within the lineage of the info asset, as proven within the following screenshot.

The enterprise steward assigned to the asset receives an approval notification.
1. Select Duties button within the high proper nook.
2. Find the newest activity titled Settle for or Reject, which is related to the aws_orders asset.

The enterprise steward opens the duty and chooses both Approve or Reject, relying on the request. On this instance, Approve is chosen. The duty is then marked as full.

After the enterprise steward approves the subscription request, the corresponding Subscription Request asset is mechanically up to date to the standing Accepted.

The requester is notified that the subscription request has been accepted. To acknowledge, the requester select Duties, locates the approval notification, and chooses Executed to substantiate receipt, as proven within the following screenshot.

After a subscription request is accepted, the mixing answer mechanically course of the request by creating and granting the corresponding subscription in AWS utilizing the asset’s metadata. The person can then verify the brand new subscription is mirrored in Amazon SageMaker, as proven within the following screenshot.

After the subscription is granted, the standing of the Subscription Request is up to date to Entry Granted.

The requester now receives a brand new activity, which is a notification confirming that the subscription request has been granted. Select the Ship button to acknowledge and full the duty.

Within the AWS Subscription Requests area, all requests and their standing are seen. Along with Accepted and Entry Granted statuses, Rejected requests are additionally listed. If a request is rejected by the approver, its standing modifications to Rejected and no subscription is created in AWS.

Synchronization Interval

The answer retains Collibra and Amazon SageMaker Catalog in sync by common updates. Core components together with enterprise metadata of Collibra, person profiles, venture data & revealed property of Amazon SageMaker Catalog, and subscription requests originating in Collibra are synchronized each 5 minutes. Nevertheless, when subscription requests are created in Amazon SageMaker Catalog, they’re immediately synchronized to Collibra.

Cleanup

To keep away from incurring pointless prices after testing or exploring the answer, delete the provisioned sources. Comply with these steps:

Take away the CloudFormation stack – Go to the AWS CloudFormation console, choose the stack you created for this answer, and select Delete. This can mechanically take away the related AWS sources provisioned by the stack.
Clear up Collibra configurations – Within the Collibra setting, take away check domains, tasks, or workflows created for this answer to make sure a clear slate for future experiments.
Revoke entry tokens or credentials – Should you used API credentials or entry tokens for integration, guarantee they’re revoked or deleted if now not wanted.

Performing these steps ensures your environments keep clear and also you keep away from unintended useful resource utilization.

Conclusion

The answer connecting Amazon SageMaker Catalog and Collibra provides organizations a easy solution to unify metadata and streamline entry workflows. It helps scale back duplication, enhance governance, and construct belief in information for each analytics and AI.We demonstrated methods to synchronize metadata and handle entry requests utilizing APIs, enabling a shared view of information throughout groups.Be taught extra by exploring:

We welcome your suggestions as you discover what’s attainable with this answer.

Concerning the authors

Vasiliki Nikolopoulou is a Principal Integrations Architect at Collibra, the place she is working for the previous 11 years. Her in depth profession consists of roles akin to Director, Enterprise Architect at AXA Insurance coverage US, Principal Gross sales Engineer at Oracle, and Licensed Senior IT Skilled in technical gross sales at IBM for over 15 years. She holds quite a few technical certifications. Join together with her on LinkedIn.

Divij Bhatia is a Software program Growth Engineer at AWS. He’s keen about constructing resilient and scalable cloud-native options that resolve real-world issues for patrons. His free time typically takes him outside, touring and taking pictures landscapes. Join with him on LinkedIn.

Leonardo Gomez is a Principal Analytics Specialist Options Architect at AWS. He has over a decade of expertise in information administration, serving to prospects across the globe handle their enterprise and technical wants. Join with him on LinkedIn.