DEV Community: Binath Perera

How to Develop an AI Agent for Your Software Systems Using MCP & Strands Agents SDK

Binath Perera — Sat, 05 Sep 2026 18:05:56 +0000

The AI era is a huge step above prior technology. But let's clear something up: AI isn't here to replace our existing systems; it's here to complement them.

Why do I say this? because no matter how good the agents become, there are always going to be tasks that are easier and more efficient to do in the traditional way. You know what they say.

Don't buy a Ferrari to drive to the mailbox.

Not every solution needs AI and to the ones that benefit from it, AI works best as an augmentation rather than a wholesale replacement.

That being said, you must also realize that if you do not adapt your systems to be AI consumable, you will be missing out on the Ferrari.

There are several frameworks out there for developing AI agents, but today I want to share with you on how to bridge the gap between traditional apps and AI apps using the Strands Agents SDK by AWS and the Model Context Protocol(MCP).

The Architecture: Connecting Agents to Existing Apps

I'm going to share with you how I developed a inventory operations manager agent to support small business owners.
The application consists of a REACT.js frontend and a SpringBoot backend with functionality served at the '/api' endpoints. For an agent to use this application we need what you call a MCP sever. MCP is a protocol that was introduced by Anthropic to give agents a standardized form to perform efficient operations on systems. The MCP server does not use AI. It is simply a JSON RPC endpoint served at '/mcp' allowing the agent to retrieve the following,

Tools: The specific actions the agent is authorized to execute.
Resources: The context and data that can be retrieved.
Prompts: Guidelines to go with tools and resources.

The Dual-Server Strategy: Reading vs. Writing

When building tools for agent consumption, you quickly realize that creating a custom tool for every possible data request is a bottleneck. If you want the user to be able to access the data and perform analytics on it freely, routing everything through custom application endpoints is inefficient.

Instead, I highly recommend a dual-server approach.

The Database MCP (For Free Exploration & Reads)
Connect your agent directly to an MCP server provided by your database vendor. I used the MongoDB MCP Server since I stored the data in MongoDB. This equips the agent with all the necessary tools to freely query and explore the data. Crucially, you must provide strict context boundaries to the agent to ensure it only retrieves authorized data(e.g- tenant information, access roles of the user). This ensures flexible analytics without tying you down to hardcoded app tools. The Strands SDK provides methods to easily add context for the agent.
The Application MCP(For Controlled Writes)
While letting an agent directly query a database is great for read operations, it is incredibly risky for data mutation. For operations that alter data, you must ensure the agent adheres to the business logic already written in you application.

By explicitly instructing the agent to use the Database MCP for exploration and the Application MCP for altering data, you get the best of both worlds: infinite analytical flexibility and ironclad data integrity.

Implementing with the AWS Strands Agents SDK

Writing this all up is surprisingly straightforward using the Strands Agents SDK. The SDK provides everything you need to easily add clients for your MCP servers. You can find my code here

Connecting to Amazon Bedrock models require almost zero configuration. It's just a matter of exposing your AWS credentials(Your access key id and access key secret) in the environment. By default the SDK will seamlessly hook into Anthropic's Claude Sonnet via Bedrock.

And that's it. You have a fully functional AI assistant to support your business activities.

Looking Forward

The evolution of these protocols is happening fast. The rise of WebMCP brings even more functionality to the table, allowing agents to further guide and support users.

AI isn't stripping away our traditional architectures; it's giving us standardized ways to make them infinitely smarter. It is a very exciting future.

Evolution Informed Design (A Moral Code for Developing Solutions)

Binath Perera — Mon, 25 May 2026 09:20:12 +0000

Recently I researched on the effects that environments have on the evolution of a species. As it turns out, these environments have a direct effect on their evolution. When we speak of our environments, we mostly encounter those that have been altered by humans. Humans grew large brains due to the invention of fire which allowed us to have more nutritious diets and easier digestion allowing the development of a larger brain than other animals. The term used to describe this phenomenon is artificial selection. You might have already heard the term natural selection which is natural evolution of a certain species without man made impact. But in artificial selection, human activities determine the evolution of that species.

Effects of artificial selection can be seen in plants.

Wild Mustard (Brassica oleracea): This single wild weed was selectively bred into vastly different vegetables by targeting specific plant parts --Kale: Cultivated by focusing on large, leafy growth. --Broccoli&Cauliflower: Bred by suppressing flower development and altering buds. --Cabbage: Selected for short stems and tight, dense leaves.

-Corn (Maize): Modern sweet corn was bred from a wild, inedible grass called teosinte, which originally had tiny kernels and hard shells

Effects of artificial selection can be seen in animals as well.

Dog Breeds: All domesticated dogs originated from a common wolf ancestor. Humans bred them for specific tasks and appearances, ranging from tiny toy poodles to massive work dogs.
Dairy Cattle: Farmers systematically breed cows that produce higher milk yields with bulls from similar high-producing lineages, drastically increasing global dairy production.

Artificial selection is not always positive.

Bulldogs have been bred for their distinctive flat faces, but this can cause breathing difficulties and other health issues.
Many modern wheat varieties are unable to effectively disperse their seeds without human help. This means they are less able to survive in the wild, making them more dependent on farmers for their continued existence.

The above examples are for plants and animals. But this applies to humans too. How about the environments that we develop like the cities and our lifestyle choices? These will have direct influence. Too much reliance on technology might shrink human brain cognitive capabilities for example. Use of antibiotics develops more resistant bacteria. Using medicine as a substitute to addressing the cause of ailment is another. Check out this video by this MIT educated former neurosurgeon https://www.youtube.com/watch?v=25LUF8GmbFU

This has to do with capitalism as well because you cannot make money by removing the cause for the problem but you can make money by providing a solution to it. And don't get me wrong, providing a solution will fix the problem, and is often necessary, but it could be a temporary fix because the solution did not address the root cause for the problem. This doesn't mean that humans should get rid of all the root causes which is anger, greed and ignorance as mentioned by Lord Buddha and become saints because that is not possible for most people. But we must collectively as a species establish standards and have a sustainable lifestyles. Solutions and inventions are great, but we must stop making solutions to problems that we have created! Because these solutions are often used as a justification to not stop the cause for the problem, thereby creating more and more complications. Let me give you a few more examples.

Plastic Recycling Campaigns : An explosion of single-use, non-biodegradable packaging that chokes oceans and landfills. By shifting the responsibility of managing waste onto the consumer ("Just put it in the blue bin!"), manufacturers successfully lobbied against bans on single-use plastics. The existence of recycling justified infinite production. Only about 9% of all plastic ever made has actually been recycled. The rest ends up in landfills, incinerators, or the environment, breaking down into microplastics that are now found in human blood and rainwater. A group of young engineers developed a novel method of extracting microplastic from seawater.
Carbon Capture and Storage : Massive greenhouse gas emissions from burning fossil fuels. Developing expensive technology to suck carbon dioxide out of the air or trap it at smokestacks and bury it underground. Fossil fuel companies heavily promote and invest in CCS because it validates their core business model. The narrative becomes, "We don't need to stop extracting and burning oil and gas, because we will eventually just vacuum up the pollution." CCS is highly energy-intensive, incredibly expensive, and has repeatedly failed to meet capture targets. Relying on it delays the transition to renewable energy, while the continued extraction of fossil fuels still destroys local ecosystems and pollutes air and water in ways CCS cannot fix.
Medicalizing the Modern Diet: Epidemic levels of obesity, type 2 diabetes, and heart disease driven by a food system optimized for cheap, hyper-palatable, ultra-processed foods. Blockbuster pharmaceutical interventions, from statins to modern GLP-1 weight-loss drugs like Ozempic and Wegovy. The availability of pharmaceutical safety nets allows the food and agriculture industries to continue marketing highly subsidized, nutrient-poor food without facing severe regulatory crackdowns (like sugar taxes, marketing bans to children, or subsidy reforms). Millions of people are effectively tethered to expensive, lifelong medications with potential side effects, simply to counteract the food environment they live in. It drains healthcare systems financially while ignoring the root cause of widespread agricultural and nutritional failure.

Artificial selection driven by human preferences or economic utility rather than an organism’s natural fitness frequently introduces harmful mutations, reduces genetic diversity, and produces traits that would be fatal in the wild.

As with any society, individuals alone cannot influence this. Take for example circular economy that clearly requires government intervention to succeed. The government must step in to oversee evolution informed design as well. Circular economy itself is a hard problem to solve, and evolution informed design is even harder. But it's crucial we think about it.

The purpose for writing this article is two fold. First is to share the important message that we have to place more importance on addressing the root cause for the problem in addition to just providing an solution to it (Buddhist philosophy teaches this btw, have a look). The second is to consider the evolutionary impact before developing and deploying whatever the solutions out to the world over narrow capitalist gains fueled by greed and ignorance. Governments must promote these ideas within society, it will help towards a more sustainable world.

From a consumer point of view, be smart. Live a healthy lifestyle. You have a choice. Give your support on making the policies that bring actual progress. I strongly believe that capitalism is the best economic philosophy we have, but it will only work out in the long term if you have the proper policies to govern it.

Thanks for reading.

Edge Computing - Pros, Cons and Common Mistakes

Binath Perera — Mon, 25 May 2026 07:55:33 +0000

Edge computing is simply localized data processing. But is it still edge computing if it is not connected to the internet? The answer is yes, it is still considered edge computing even when disconnected from the cloud. Edge computing is defined by processing data locally at or near the source—such as on local devices, IoT gateways, or on-premises servers—rather than sending it entirely to a centralized cloud.

But edge computing is often utilized as part of a larger system containing fog computing and cloud computing.

Let's look at the Pros and Cons of processing your data at the edge compared to central cloud processing,

Pros,

Prevents network congestion while saving valuable bandwidth
Reduced response times
Ideal for real time technologies
Reduced surface exposure to private data since the data is processed at the closest location.

Cons,

Requires more hardware components
Complex maintenance

Faults system developers make

Taking security shortcuts: Anyone who has physical access to your edge devices can exploit it.
Over Engineering
Management Nightmare: Not designing with management in mind
Not testing in real world production environment
Not managing the limited resources effectively (Power, Memory etc)

Using a combination of edge, fog and cloud computing is the future of industry 4.0 and beyond. Remembering the pros, cons and common mistakes will help to develop effective edge computing solutions for the modern world.

AWS Student Community Day Sri Lanka 2026 - Event Recap

Binath Perera — Thu, 30 Apr 2026 15:04:39 +0000

AWS Student Community Days are one day, student led conferences, organized by AWS Student Builder Groups (Formerly AWS Cloud Clubs). It is a great way to bringing an AWS conference into your university.

On 16th May 2026 Sri Lanka had it's first AWS Student Community Day at the University of Kelaniya. The event was co-organized by the student builder groups at the University of Kelaniya and Institute of Technology University of Moratuwa. The event had 14 speakers and over 200 student attendees. It was a full day of community building and learning for the students.

I as the Builder Group Leader from the University of Kelaniya created this article to impart my learnings to you, based on my experience in organizing this awesome event on campus.

The approach that me and Naami (The builder group leader from ITUM) took was we first made an organizing committee of student enthusiasts who would like to take part in organizing this event. Based on my past experience hosting events, I learnt that each member in a team needs to know their responsibilities clearly. To ensure this, a document was made laying out the responsibilities of each role before calling in for applications to the organizing committee. In this way every team member was aware of their responsibilities. Once we received the applications, the OC members were selected. The first OC meeting for the exco members (THe VPs of each team) was taken 45 days before the event date.

Our Organizing Committee structure was designed as follows.

I realized later down the line that it Would have been better to call the marketing and communications team marketing and production. I missed out on specifying that the designing team is also responsible for event day production visuals and other content, not just marketing.

The 2 leaders and other exco members took meetings each week to catch up with the tasks done by each team and made a plan for the following week.

Here are some important requirements when hosting a conference event.

Web page
This is one of the first things you are going to need. It's up to you whether you create your own or use some template. We used a template from AWS Community Day China and asked lovable to generate a similar one. Then we customized it to our needs.

Registation
We used to local ticketing platform called MyTickets to track registrations and generate QR codes for attendees. Yes we did issue tickets. The community day organizers are always dealing with this one... and there is no right or wrong way. Both have pros and cons.

Paid Event - Even symbolic price can reduce the no-shows (ratio between registered and the ones that actually showed-up) and increase the budget you get. But there is a chance you have to pay taxes, as you are creating the profit.

Free Event - Prepare yourself for a no-shows... 😬 It's frustrating, but it is what is is.

We decided to go with paid.

Call for speakers and sponsors
We used two google forms, one to call for speakers and the other to call for sponsors. We also crafted a sponsorship prospectus and attached it in our call for sponsors form.

The sponsorship team worked hard in securing sponsorships and we found one sponsor, WSO2 and SLASSCOM as the ecosystem partner. You should reach out to companies way before the event. Do not wait until the last moment. We were lucky to get those partners at the last moment, but could have gotten more if we planed it well in advance.

The event was a great success, 14 speakers and over 200 students, connecting over cloud computing technology. It was a fun and educational day for the students. The support from the AWS community partners, AWS User Group Colombo and the community builders have been tremendous to make the event a success.

All being said, organizing AWS Student Community Day was a lot of fun, but also lots of hard work went into it. It took us 3 months of work, from idea that we are doing that, to the actual event.

If you are still thinking if to do it or not host such an event on campus, what are you waiting for?, go for it! 😉