Big Data

Amazon DataZone introduces OpenLineage-compatible information lineage visualization in preview

09/07/2024
florenc85

[ad_1]

We’re excited to announce the preview of API-driven, OpenLineage-compatible information lineage in Amazon DataZone that will help you seize, retailer, and visualize lineage of knowledge motion and transformations of knowledge property on Amazon DataZone.

With the Amazon DataZone OpenLineage-compatible API, area directors and information producers can seize and retailer lineage occasions past what is on the market in Amazon DataZone, together with transformations in Amazon Easy Storage Service (Amazon S3), AWS Glue, and different AWS providers. This supplies a complete view for information customers shopping in Amazon DataZone, who can acquire confidence of an asset’s origin, and information producers, who can assess the influence of adjustments to an asset by understanding its utilization.

On this put up, we focus on the newest options of knowledge lineage in Amazon DataZone, its compatibility with OpenLineage, and the best way to get began capturing lineage from different providers akin to AWS Glue, Amazon Redshift, and Amazon Managed Workflows for Apache Airflow (Amazon MWAA) into Amazon DataZone by the API.

Why it issues to have information lineage

Knowledge lineage provides you an overarching view into information property, permitting you to see the origin of objects and their chain of connections. Knowledge lineage allows monitoring the motion of knowledge over time, offering a transparent understanding of the place the information originated, the way it has modified, and its final vacation spot throughout the information pipeline. With transparency round information origination, information customers acquire belief that the information is right for his or her use case. Knowledge lineage data is captured at ranges akin to tables, columns, and jobs, permitting you to conduct influence evaluation and reply to information points as a result of, for instance, you’ll be able to see how one discipline impacts downstream sources. This equips you to make well-informed choices earlier than committing adjustments and keep away from undesirable adjustments downstream.

Knowledge lineage in Amazon DataZone is an API-driven, OpenLineage-compatible characteristic that helps you seize and visualize lineage occasions from OpenLineage-enabled methods or by an API, to hint information origins, monitor transformations, and look at cross-organizational information consumption. The lineage visualized contains actions contained in the Amazon DataZone enterprise information catalog. Lineage captures the property cataloged in addition to the subscribers to these property and to actions that occur outdoors the enterprise information catalog captured programmatically utilizing the API.

Moreover, Amazon DataZone variations lineage with every occasion, enabling you to visualise lineage at any cut-off date or examine transformations throughout an asset’s or job’s historical past. This historic lineage supplies a deeper understanding of how information has advanced, which is crucial for troubleshooting, auditing, and implementing the integrity of knowledge property.

The next screenshot exhibits an instance lineage graph visualized with the Amazon DataZone information catalog.

Introduction to OpenLineage appropriate information lineage

The necessity to seize information lineage persistently throughout varied analytical providers and mix them right into a unified object mannequin is essential in uncovering insights from the lineage artifact. OpenLineage is an open supply challenge that gives a framework to gather and analyze lineage. It additionally affords reference implementation of an object mannequin to persist metadata together with integration to main information and analytics instruments.

The next are key ideas in OpenLineage:

  • Lineage occasions – OpenLineage captures lineage data by a sequence of occasions. An occasion is something that represents a particular operation carried out on the information that happens in a knowledge pipeline, akin to information ingestion, transformation, or information consumption.
  • Lineage entities – Entities in OpenLineage characterize the assorted information objects concerned within the lineage course of, akin to datasets and tables.
  • Lineage runs – A lineage run represents a particular run of a knowledge pipeline or a job, encompassing a number of lineage occasions and entities.
  • Lineage kind varieties – Kind varieties, or aspects, present extra metadata or context about lineage entities or occasions, enabling richer and extra descriptive lineage data. OpenLineage affords aspects for runs, jobs, and datasets, with the choice to construct customized aspects.

The Amazon DataZone information lineage API is OpenLineage appropriate and extends OpenLineage’s performance by offering a materialization endpoint to persist the lineage outputs in an extensible object mannequin. OpenLineage affords integrations for sure sources, and integration of those sources with Amazon DataZone is simple as a result of the Amazon DataZone information lineage API understands the format and interprets to the lineage information mannequin.

The next diagram illustrates an instance of the Amazon DataZone lineage information mannequin.

In Amazon DataZone, each lineage node represents an underlying useful resource—there’s a 1:1 mapping of the lineage node with a logical or bodily useful resource akin to desk, view, or asset. The nodes characterize a particular job with a particular run, or a node for a desk or asset, and one node for a subscription goal.

Every model of a node captures what occurred to the underlying useful resource at that particular timestamp. In Amazon DataZone, lineage not solely shares the story of knowledge motion outdoors it, however it additionally represents the lineage of actions inside Amazon DataZone, akin to asset creation, curation, publishing, and subscription.

To hydrate the lineage mannequin in Amazon DataZone, two varieties of lineage are captured:

  • Lineage actions inside Amazon DataZone – This contains property added to the catalog and printed, after which particulars concerning the subscriptions are captured robotically. If you’re within the producer challenge context (for instance, if the challenge you’re chosen is the proudly owning challenge of the asset you might be shopping and also you’re a member of that challenge), you will notice two states of the dataset node:
    • The stock asset kind node defines the asset within the catalog that’s in an unpublished stage. Different customers can’t subscribe to the stock asset. To be taught extra, discuss with Creating stock and printed information in Amazon DataZone.
    • The printed asset kind represents the precise asset that’s discoverable by information customers throughout the group. That is the asset kind that may be subscribed by different challenge members. If you’re a shopper and never a part of the manufacturing challenge of that asset, you’ll solely see the printed asset node.
  • Lineage actions outdoors of Amazon DataZone will be captured programmatically utilizing the PostLineageEvent With these occasions captured both upstream or downstream of cataloged property, information producers and customers get a complete view of knowledge motion to test the origin of knowledge or its consumption. We focus on the best way to use the API to seize lineage occasions later on this put up.

There are two several types of lineage nodes obtainable in Amazon DataZone:

  • Dataset node – In Amazon DataZone, lineage visualizes nodes that characterize tables and views. Relying on the context of the challenge, the producers will be capable of view each the stock and printed asset, whereas customers can solely view the printed asset. If you first open the lineage tab on the asset particulars web page, the cataloged dataset node would be the start line for lineage graph traversal upstream or downstream. Dataset nodes embrace lineage nodes automated from Amazon DataZone and customized lineage nodes:
    • Automated dataset nodes – These nodes embrace details about AWS Glue or Amazon Redshift property printed within the Amazon DataZone catalog. They’re robotically generated and embrace a corresponding AWS Glue or Amazon Redshift icon throughout the node.
    • Customized dataset nodes – These nodes embrace details about property that aren’t printed within the Amazon DataZone catalog. They’re created manually by area directors (producers) and are represented by a default customized asset icon throughout the node. These are basically customized lineage nodes created utilizing the OpenLineage occasion format.
  • Job (run) node – This node captures the small print of the job, which represents the newest run of a specific job and its run particulars. This node additionally captures a number of runs of the job and will be considered on the Historical past tab of the node particulars. Node particulars are made seen while you select the icon.

Visualizing lineage in Amazon DataZone

Amazon DataZone affords a complete expertise for information producers and customers. The asset particulars web page supplies a graphical illustration of lineage, making it easy to visualise information relationships upstream or downstream. The asset particulars web page supplies the next capabilities to navigate the graph:

  • Column-level lineage – You may increase column-level lineage when obtainable in dataset nodes. This robotically exhibits relationships with upstream or downstream dataset nodes if supply column data is on the market.
  • Column search – If the dataset has greater than 10 columns, the node presents pagination to navigate to columns not initially offered. To rapidly view a specific column, you’ll be able to search on the dataset node that lists simply the searched column.
  • View dataset nodes solely – If you would like filter out the job nodes, you’ll be able to select the Open view management icon within the graph viewer and toggle the Show dataset nodes solely This can take away all of the job nodes from the graph and allow you to navigate simply the dataset nodes.
  • Particulars pane – Every lineage node captures and shows the next particulars:
    • Each dataset node has three tabs: Lineage data, Schema, and Historical past. The Historical past tab lists the totally different variations of lineage occasion captured for that node.
    • The job node has a particulars pane to show job particulars with the tabs Job data and Historical past. The main points pane additionally captures queries or expressions run as a part of the job.
  • Model tabs – All lineage nodes in Amazon DataZone information lineage may have versioning, captured as historical past, primarily based on lineage occasions captured. You may view lineage at a particular timestamp that opens a brand new tab on the lineage web page to assist examine or distinction between the totally different timestamps.

The next screenshot exhibits an instance of knowledge lineage visualization.

You may expertise the visualization with pattern information by selecting Preview on the Lineage tab and selecting the Attempt pattern lineage hyperlink. This opens a brand new browser tab with pattern information to check and be taught concerning the characteristic with or with out a guided tour, as proven within the following screenshot.

Answer overview

Now that we perceive the capabilities of the brand new information lineage characteristic in Amazon DataZone, let’s discover how one can get began in capturing lineage from AWS Glue tables and ETL (extract, rework, and cargo) jobs, Amazon Redshift, and Amazon MWAA.

The getting began scripts are additionally obtainable in Amazon DataZone’s new GitHub repository.

Conditions

For this walkthrough, it’s best to have the next conditions:

If the AWS account you utilize to observe this put up makes use of AWS Lake Formation to handle permissions on the AWS Glue Knowledge Catalog, just remember to log in as a person with entry to create databases and tables. For extra data, discuss with Implicit Lake Formation permissions.

Launch the CloudFormation stack

To create your sources for this use case utilizing AWS CloudFormation, full the next steps:

  1. Launch the CloudFormation stack in us-east-1:
  2. For Stack title, enter a reputation to your stack.
  3. Select Subsequent.
  4. Choose I acknowledge that AWS CloudFormation would possibly create IAM sources with customized names.
  5. Select Create stack.

Look ahead to the stack formation to complete provisioning the sources. If you see the CREATE_COMPLETE standing, you’ll be able to proceed to the following steps.

Seize lineage from AWS Glue tables

For this instance, we use CloudShell, which is a browser-based shell, to run the instructions needed to reap lineage metadata from AWS Glue tables. Full the next steps:

  1. On the AWS Glue console, select Crawlers within the navigation pane.
  2. Choose the AWSomeRetailCrawler crawler created by the CloudFormation template.
  3. Select Run.

When the crawler is full, you’ll see a Succeeded standing.

Now let’s harvest the lineage metadata utilizing CloudShell.

  1. Obtain the extract_glue_crawler_lineage.py file.
  2. On the Amazon DataZone console, open CloudShell.
  1. On the Actions menu, select Replace file.
  2. Add the extract_glue_crawler_lineage.py file.

  3. Run the next instructions: 
    sudo yum -y set up python3
python3 -m venv env
. env/bin/activate
pip set up boto3

It is best to get the next outcomes.

  1. After all of the libraries and dependencies are configured, run the next command to reap the lineage metadata from the stock desk: 
    python extract_glue_crawler_lineage.py -d awsome_retail_db -t stock -r us-east-1 -i dzd_Your_doamin

  2. The script asks for verification of the settings offered; enter Sure.

It is best to obtain a notification indicating that the script ran efficiently.

After you seize the lineage data from the Stock desk, full the next steps to run the information supply.

  1. On the Amazon DataZone information portal, open the Gross sales
  2. On the Knowledge tab, select Knowledge sources within the navigation pane.
  1. Choose your information supply job and select Run.

For this instance, we had a knowledge supply job known as SalesDLDataSourceV2 already created pointing to the awesome_retail_db database. To be taught extra about the best way to create information supply jobs, discuss with Create and run an Amazon DataZone information supply for the AWS Glue Knowledge Catalog.

After the job runs efficiently, it’s best to see a affirmation message.

Now let’s view the lineage diagram generated by Amazon DataZone.

  1. On the Knowledge stock tab, select the Stock desk.
  2. On the Stock asset web page, select the brand new Lineage tab.

On the Lineage tab, you’ll be able to see that Amazon DataZone created three nodes:

  • Job / Job run – That is primarily based on the AWS Glue crawler used to reap the asset technical metadata
  • Dataset – That is primarily based on the S3 object that accommodates the information associated to this asset
  • Desk – That is the AWS Glue desk created by the crawler

If you happen to select the Dataset node, Amazon DataZone affords details about the S3 object used to create the asset.

Seize information lineage for AWS Glue ETL jobs

Within the earlier part, we lined the best way to generate a knowledge lineage diagram on high of a knowledge asset. Now let’s see how we are able to create one for an AWS Glue job.

The CloudFormation template that we launched earlier created an AWS Glue job known as Inventory_Insights. This job will get information from the Stock desk and creates a brand new desk known as Inventory_Insights with the aggregated information of the full merchandise obtainable in all of the shops.

The CloudFormation template additionally copied the openlineage-spark_2.12-1.9.1.jar file to the S3 bucket created for this put up. This file is critical to generate lineage metadata from the AWS Glue job. We use model 1.9.1, which is appropriate with AWS Glue 3.0, the model used to create the AWS Glue job for this put up. If you happen to’re utilizing a distinct model of AWS Glue, you must obtain the corresponding OpenLineage Spark plugin file that matches your AWS Glue model.

The OpenLineage Spark plugin will not be in a position to extract information lineage from AWS Glue Spark jobs that use AWS Glue DynamicFrames. Use Spark SQL DataFrames as a substitute.

  1. Obtain the extract_glue_spark_lineage.py file.
  2. On the Amazon DataZone console, open CloudShell.
  3. On the Actions menu, select Replace file.
  4. Add the extract_glue_spark_lineage.py file.
  5. On the CloudShell console, run the next command (in case your CloudShell session expired, you’ll be able to open a brand new session): 
    python extract_glue_spark_lineage.py —area "us-east-1" —domain-identifier 'dzd_Your Area'

  6. Affirm the knowledge confirmed by the script by coming into sure.

You will notice the next message; which means the script is able to get the AWS Glue job lineage metadata after you run it.

Now let’s run the AWS Glue job created by the Cloud formation template.

  1. On the AWS Glue console, select ETL jobs within the navigation pane.
  2. Choose the Inventory_Insights job and select Run job.

On the Job particulars tab, you’ll discover that the job has the next configuration:

  • Key --conf with worth extraListeners=io.openlineage.spark.agent.OpenLineageSparkListener --conf spark.openlineage.transport.kind=console --conf spark.openlineage.aspects.custom_environment_variables=[AWS_DEFAULT_REGION;GLUE_VERSION;GLUE_COMMAND_CRITERIA;GLUE_PYTHON_VERSION;]
  • Key --user-jars-first with worth true
  • Dependent JARs path set because the S3 path s3://{your bucket}/lib/openlineage-spark_2.12-1.9.1.jar
  • The AWS Glue model set as 3.0

Throughout the run of the job, you will notice the next output on the CloudShell console.

Which means that the script has efficiently harvested the lineage metadata from the AWS Glue job.

Now let’s create an AWS Glue desk primarily based on the information created by the AWS Glue job. For this instance, we use an AWS Glue crawler.

  1. On the AWS Glue console, select Crawlers within the navigation pane.
  2. Choose the AWSomeRetailCrawler crawler created by the CloudFormation template and select Run.

When the crawler is full, you will notice the next message.

Now let’s open the Amazon DataZone portal to see how the diagram is represented in Amazon DataZone.

  1. On the Amazon DataZone portal, select the Gross sales challenge.
  2. On the Knowledge tab, select Stock information within the navigation pane.
  3. Select the stock insights asset

On the Lineage tab, you’ll be able to see the diagram created by Amazon DataZone. It exhibits three nodes:

    • The AWS Glue crawler used to create the AWS Glue desk
    • The AWS Glue desk created by the crawler
    • The Amazon DataZone cataloged asset
  1. To see the lineage details about the AWS Glue job that you simply ran to create the inventory_insights desk, select the arrows icon on the left facet of the diagram.

Now you’ll be able to see the complete lineage diagram for the Inventory_insights desk.

  1. Select the blue arrow icon within the stock node to the left of the diagram.

You may see the evolution of the columns and the transformations that that they had.

If you select any of the nodes which might be a part of the diagram, you’ll be able to see extra particulars. For instance, the inventory_insights node exhibits the next data.

Seize lineage from Amazon Redshift

Let’s discover the best way to generate a lineage diagram from Amazon Redshift. On this instance, we use AWS Cloud9 as a result of it permits us to configure the connection to the digital non-public cloud (VPC) the place our Redshift cluster resides. For extra details about AWS Cloud9, discuss with the AWS Cloud9 Consumer Information.

The CloudFormation template included as a part of this put up doesn’t cowl the creation of a Redshift cluster or the creation of the tables used on this part. To be taught extra about the best way to create a Redshift cluster, see Step 1: Create a pattern Amazon Redshift cluster. We use the next question to create the tables wanted for this part of the put up:

Create SCHEMA market

create desk market.retail_sales (
  id BIGINT major key,
  title character various not null
);

create desk market.online_sales (
  id BIGINT major key,
  title character various not null
);

/* Essential to insert some information within the desk */
INSERT INTO market.retail_sales
VALUES (123, 'item1')

INSERT INTO market.online_sales
VALUES (234, 'item2')

create desk market.gross sales AS
Choose id, title from market.retail_sales
Union ALL
Choose id, title from market.online_sales;

Bear in mind so as to add the IP handle of your AWS Cloud9 setting to the safety group with entry to the Redshift cluster.

  1. Obtain the necessities.txt and extract_redshift_lineage.py recordsdata.
  2. On the File menu, select Add Native Information.
  3. Add the necessities.txt and extract_redshift_lineage.py recordsdata.
  4. Run the next instructions: 
    # Set up Python 
sudo yum -y set up python3

# dependency arrange 
python3 -m venv env 
. env/bin/activate

pip set up -r necessities.txt

It is best to be capable of see the next messages.

  1. To set the AWS credentials, run the next command: 
    export AWS_ACCESS_KEY_ID=<<Your Entry Key>>
export AWS_SECRET_ACCESS_KEY=<<Your Secret Entry Key>>
export AWS_SESSION_TOKEN=<<Your Session Token>>

  2. Run the extract_redshift_lineage.py script to reap the metadata essential to generate the lineage diagram: 
    python extract_redshift_lineage.py 
 -r area 
 -i dzd_your_dz_domain_id 
 -n your-redshift-cluster-endpoint 
 -t your-rs-port 
 -d your-database 
 -s the-starting-date

  3. Subsequent, you’ll be prompted to enter the person title and password for the connection to your Amazon DataZone database.
  4. If you obtain a affirmation message, enter sure.

If the configuration was performed accurately, you will notice the next affirmation message.

Now let’s see how the diagram was created in Amazon DataZone.

  1. On the Amazon DataZone information portal, open the Gross sales challenge.
  2. On the Knowledge tab, select Knowledge sources.
  3. Run the information supply job.

For this put up, we already created a knowledge supply job known as Sales_DW_Enviroment-default-datasource so as to add the Redshift information supply to our Amazon DataZone challenge. To discover ways to create a knowledge supply job, discuss with Create and run an Amazon DataZone information supply for Amazon Redshift

After you run the job, you’ll see the next affirmation message.

  1. On the Knowledge tab, select Stock information within the navigation pane.
  2. Select the total_sales asset.
  1. Select the Lineage tab.

Amazon DataZone create a three-node lineage diagram for the full gross sales desk; you’ll be able to select any node to view its particulars.

  1. Select the arrows icon subsequent to the Job/ Job run node to view a extra full lineage diagram.
  1. Select the Job / Job run

The Job Information part exhibits the question that was used to create the full gross sales desk.

Seize lineage from Amazon MWAA

Apache Airflow is an open-source platform for growing, scheduling, and monitoring batch-oriented workflows. Amazon MWAA is a managed service for Airflow that permits you to use your present Airflow platform to orchestrate your workflows. OpenLineage helps integration with Airflow 2.6.3 utilizing the openlineage-airflow package deal, and the identical will be enabled on Amazon MWAA as a plugin. As soon as enabled, the plugin converts Airflow metadata to OpenLineage occasions, that are consumable by DataZone.PostLineageEvent.

The next diagram exhibits the setup required in Amazon MWAA to seize information lineage utilizing OpenLineage and publish it to Amazon DataZone.

The workflow makes use of an Amazon MWAA DAG to invoke a knowledge pipeline. The method is as follows:

  1. The openlineage-airflow plugin is configured on Amazon MWAA as a lineage backend. Metadata concerning the DAG run is handed to the plugin, which converts it into OpenLineage format.
  2. The lineage data collected is written to Amazon CloudWatch log group based on the Amazon MWAA setting.
  3. A helper operate captures the lineage data from the log file and publishes it to Amazon DataZone utilizing the PostLineageEvent API.

The instance used within the put up makes use of Amazon MWAA model 2.6.3 and OpenLineage plugin model 1.4.1. For different Airflow variations supported by OpenLineage, discuss with Supported Airflow variations.

Configure the OpenLineage plugin on Amazon MWAA to seize lineage

When harvesting lineage utilizing OpenLineage, a Transport configuration must be arrange, which tells OpenLineage the place to emit the occasions to, for instance the console or an HTTP endpoint. You should use ConsoleTransport, which logs the OpenLineage occasions within the Amazon MWAA activity CloudWatch log group, which might then be printed to Amazon DataZone utilizing a helper operate.

Specify the next within the necessities.txt file added to the S3 bucket configured for Amazon MWAA:

openlineage-airflow==1.4.1

Within the Airflow logging configuration part beneath the MWAA configuration for the Airflow setting, allow Airflow activity logs with log stage INFO. The next screenshot exhibits a pattern configuration.

A profitable configuration will add a plugin to Airflow, which will be verified from the Airflow UI by selecting Plugins on the Admin menu.

On this put up, we use a pattern DAG to hydrate information to Redshift tables. The next screenshot exhibits the DAG in graph view.

Run the DAG and upon profitable completion of a run, open the Amazon MWAA activity CloudWatch log group to your Airflow setting (airflow-env_name-task) and filter primarily based on the expression console.py to pick occasions emitted by OpenLineage. The next screenshot exhibits the outcomes.

Publish lineage to Amazon DataZone

Now that you’ve the lineage occasions emitted to CloudWatch, the following step is to publish them to Amazon DataZone to affiliate them to a knowledge asset and visualize them on the enterprise information catalog.

  1. Obtain the recordsdata necessities.txt and airflow_cw_parse_log.py and collect setting particulars like AWS area, Amazon MWAA setting title and Amazon DataZone Area ID.
  2. The Amazon MWAA setting title will be obtained from the Amazon MWAA console.
  3. The Amazon DataZone area ID will be obtained from Amazon DataZone service console or from the Amazon DataZone portal.
  4. Navigate to CloudShell and select Add recordsdata on the Actions menu to add the recordsdata necessities.txt and extract_airflow_lineage.py.

  5. After the recordsdata are uploaded, run the next script to filter lineage occasions from the Airflow activity logs and publish them to Amazon DataZone: 
    # Arrange digital env and set up dependencies
python -m venv env
pip set up -r necessities.txt
. env/bin/activate

# run the script
python extract_airflow_lineage.py 
  --region us-east-1 
  --domain-identifier your_domain_identifier 
  --airflow-environment-name your_airflow_environment_name

The operate extract_airflow_lineage.py filters the lineage occasions from the Amazon MWAA activity log group and publishes the lineage to the desired area inside Amazon DataZone.

Visualize lineage on Amazon DataZone

After the lineage is printed to DataZone, open your DataZone challenge, navigate to the Knowledge tab and selected a knowledge asset that was accessed by the Amazon MWAA DAG. On this case, it’s a subscribed asset.

Navigate to the Lineage tab to visualise the lineage printed to Amazon DataZone.

Select a node to take a look at extra lineage metadata. Within the following screenshot, we are able to observe the producer of the lineage has been marked as airflow.

Conclusion

On this put up, we shared the preview characteristic of knowledge lineage in Amazon DataZone, the way it works, and how one can seize lineage occasions, from AWS Glue, Amazon Redshift, and Amazon MWAA, to be visualized as a part of the asset shopping expertise.

To be taught extra about Amazon DataZone and the best way to get began, discuss with the Getting began information. Take a look at the YouTube playlist for a few of the newest demos of Amazon DataZone and quick descriptions of the capabilities obtainable.

Concerning the Authors

Leonardo Gomez is a Principal Analytics Specialist at AWS, with over a decade of expertise in information administration. Specializing in information governance, he assists clients worldwide in maximizing their information’s potential whereas selling information democratization. Join with him on LinkedIn.

Priya Tiruthani is a Senior Technical Product Supervisor with Amazon DataZone at AWS. She focuses on bettering information discovery and curation required for information analytics. She is keen about constructing revolutionary merchandise to simplify clients’ end-to-end information journey, particularly round information governance and analytics. Exterior of labor, she enjoys being open air to hike, seize nature’s magnificence, and not too long ago play pickleball.

Ron Kyker is a Principal Engineer with Amazon DataZone at AWS, the place he helps drive innovation, remedy complicated issues, and set the bar for engineering excellence for his workforce. Exterior of labor, he enjoys board gaming with family and friends, films, and wine tasting.

Srinivasan Kuppusamy is a Senior Cloud Architect – Knowledge at AWS ProServe, the place he helps clients remedy their enterprise issues utilizing the facility of AWS Cloud expertise. His areas of pursuits are information and analytics, information governance, and AI/ML.

[ad_2]

Leave a Reply

Your email address will not be published. Required fields are marked *