DEV Community: Mohan Sharma

AWS Organization equivalent concept in Azure

Mohan Sharma — Wed, 18 Feb 2026 16:35:37 +0000

How to add custom CA Cert in Lambda function

Mohan Sharma — Wed, 11 Feb 2026 14:29:56 +0000

Steps

Put certificate in PEM format at the following path python/certs/xyz.pem. Directory structure is important.
Zip python directory and create lambda layer using this zip file.
Create lambda function under VPC using the below code and attach lambda layer created at step 2.
Set Environment variable in Lambda function. Key : REQUESTS_CA_BUNDLE Value: /opt/python/certs/xyz.pem
Test the lambda function.

How to create Lambda Layer in Python

Mohan Sharma — Tue, 10 Feb 2026 16:38:45 +0000

Use Lambda layers to package code and dependencies that you want to reuse across multiple functions. Layers usually contain library dependencies, a custom runtime, or configuration files.

Below are the steps to create lambda layer and test it with lambda function.

Check python version in your system with "py --version". This version and lambda runtime version MUST match. eg 3.13
mkdir lambda_layer
mkdir -p python
pip install requests -t python/
Zip python directory.
Go to AWS and create Lambda layer by uploading python.zip file.
Create lambda function with the same runtime version eg. 3.13 and attach lambda layer to it.
import module required modules in lambda_function.py handle file
Test the function by creating test event.

How to plan Agentic AI rollout

Mohan Sharma — Mon, 09 Feb 2026 17:01:20 +0000

The below are the different phases of rolling out AI Agents in an organization.

Crawl phase.

Deploy the first agent internally for a small pilot group. Focus on learning and iteration. Failures are cheap.

Walk phase

Deploy the agent to a controlled external user group. More users, more feedback, more edge cases discovered. Investment in observability and evaluation pays off.

Run phase

Scale the agent to external users with confidence. Platform capabilities enable other teams to build their own agents faster. Organizational capability compounds.

This is how organizations can go from one developer building one agent to dozens of teams building dozens of agents with consistent quality, shared infrastructure, and accelerating velocity.

Best Practices with Amazon AgentCore

Mohan Sharma — Mon, 09 Feb 2026 13:32:50 +0000

Building production-ready AI agents requires disciplined engineering practices across the entire lifecycle, includes

• Start small with a clearly defined problem
• Instrument everything from day one
• Build a deliberate tooling strategy
• Automate your evaluation
• Decompose complexity with multi-agent architectures
• Scale securely with personalization
• Combine agents with deterministic code
• Test continuously
• Build organizational capability with platform thinking

Amazon Bedrock AgentCore provides the services you need to implement these practices.

• AgentCore RUNTIME hosts agents and tools in isolated environments
• AgentCore MEMORY enables personalized interactions
• AgentCore IDENTITY AND POLICY help enforce security
• AgentCore OBSERVABILITY provides visibility
• AgentCore EVALUATIONS enables continuous quality assessment
• AgentCore GATEWAY unifies communication across agents and tools using standard protocols
• AgentCore BROWSER provides a secure, cloud-based browser that enables AI agents to interact with websites and AgentCore Code Interpreter enables AI agents to write and execute code more securely in sandbox environments.

Monitoring vs Observability ?

Mohan Sharma — Tue, 27 Jan 2026 14:10:27 +0000

Monitoring

Monitoring detects issues using metrics and alerts to detect problem based on thresholds.

Observability

Observability enables understanding of unknown issues through logs, metrics, and distributed traces to diagnose root causes in complex systems.

Logs

Log is an idempotent record of a discrete event that happened in a system at any point of time during the request life cycle.

Traces

Tracing gives the capability to monitor the fate of a request during its lifecycle across various components in a system.

Metrics

State/health of the system at specific point of time. eg CPU Usage

Agentic AI vs Generative AI

Mohan Sharma — Fri, 16 Jan 2026 14:57:26 +0000

Agentic AI emphasizes autonomy meaning the ability for AI to operate independently.
Such AI agent performs tasks without human oversight or step-by-step instructions.
For example, once given a high-level objective, an autonomous agent decides its course of action and carries it out, rather than waiting for a human to tell it each step. An agentic system chooses which actions to take and when to take them to meet its goals.
This autonomy is what gives agentic AI its name

What is an Agent in AI?

Mohan Sharma — Tue, 30 Dec 2025 05:15:15 +0000

The below are the example of Agents in real life.

Real Estate Agents
Customer Service Agents

They are specialized in specific area.

Similarly we can think of AI agents as software/program that uses models (often AI/ML), tools, and data to autonomously plan and perform tasks in pursuit of a defined goal, usually with guardrails and human oversight.

if we further break down, an agent is similar to microservices in software development. The way microservices are designed to serve specific task/audience, similarly agents are independent, specialized and designed to operate autonomously. What makes agent special is that Agents mirror how humans work.

Windows Service equivalent in Linux?

Mohan Sharma — Mon, 01 Dec 2025 07:48:29 +0000

A Windows service is the equivalent of a UNIX daemon. It is a process that provides one or more facilities to client processes.

Starting and Stopping Services

sudo systemctl list-units --type=service --state running                || List all running services
sudo systemctl start <service_name>                      || Starts a service.
sudo systemctl stop <service_name>                       || Stops a service.
sudo systemctl restart <service_name>                    || Restarts a service.
sudo systemctl enable <service_name>                     || Configures a service to start automatically at boot.
sudo systemctl disable <service_name>                    || Prevents a service from starting automatically at boot.
sudo systemctl status <service_name>                     || Shows the current status of a service.

Creating AWS Elastic Beanstalk Env through CLI commands

Mohan Sharma — Fri, 28 Nov 2025 11:04:14 +0000


## CREATE APPLICATION VERSION

aws elasticbeanstalk create-application-version \
--application-name my-app-test \
--version-label v1 \
--source-bundle S3Bucket=test-bucket-apse1,S3Key=python-app.zip

## CREATE CONFIGURATION TEMPLATE 
aws elasticbeanstalk create-configuration-template \
--application-name my-app-test \
--template-name v1 \
--solution-stack-name "64bit Amazon Linux 2023 v4.7.1 running Python 3.13"

## CREATE ENVIRONMENT
aws elasticbeanstalk create-environment \
--cname-prefix my-cname \
--application-name my-app-test \
--version-label v1 \
--template-name v1 \
--environment-name v1clone2 \
--option-settings file://c:/MohanS/NPPFiles/Options.txt \
--tags Key=Key1,Value=hr Key=Env,Value=dev Key=Key2,Value=Key2Value –debug

## TERMINATE Environment
aws elasticbeanstalk terminate-environment --environment-name v1clone1

OPTIONS.TXT FILE

[
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "IamInstanceProfile",
        "Value": "aws-elasticbeanstalk-ec2-role"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "InstanceType",
        "Value": "t3.xlarge"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "ImageId",
        "Value": "ami-0f77922ee968e2ec"
    },
    {
        "Namespace": "aws:elasticbeanstalk:environment",
        "OptionName": "ServiceRole",
        "Value": "servicerole-beanstalk"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "LaunchTemplateTagPropagationEnabled",
        "Value": "true"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "DisableIMDSv1",
        "Value": "true"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "RootVolumeType",
        "Value": "gp3"
    },
    {
        "Namespace": "aws:autoscaling:launchconfiguration",
        "OptionName": "RootVolumeSize",
        "Value": "64"
    },
    {
        "Namespace": "aws:ec2:vpc",
        "OptionName": "VPCId",
        "Value": "vpc-12345c5c68c0e840"
    },
    {
        "Namespace": "aws:ec2:vpc",
        "OptionName": "Subnets",
        "Value": "subnet-123456275dcb5,subnet-789012c147184c9"
    },
    {
        "Namespace": "aws:ec2:vpc",
        "OptionName": "ELBSubnets",
        "Value": "subnet-123456275dcb5,subnet-789012c147184c9"
    },
    {
        "Namespace": "aws:ec2:vpc",
        "OptionName": "ELBScheme",
        "Value": "internal"
    },
    {
        "Namespace": "aws:ec2:vpc",
        "OptionName": "AssociatePublicIpAddress",
        "Value": "false"
    }

]

Installation of Docker Engine in Linux EC2

Mohan Sharma — Fri, 28 Nov 2025 10:40:52 +0000

sudo su
sudo yum update -y || Apply pending updates
sudo yum search docker || SEARCH DOCKER PACKAGE
sudo yum info docker || Get version information
sudo yum install docker || Install docker
sudo docker --version || check if docker is installed.

Checking .Net Framework version installed

Mohan Sharma — Wed, 05 Nov 2025 16:07:30 +0000

PS C:\Windows\system32> Get-WindowsFeature -Name NET-Framework* | Format-Table Name, DisplayName, InstallState