DEV Community: Maryam Yusuf

How I Built MataConnect DataAgent: Connecting Women’s Communities with Multi-Agent Intelligence

Maryam Yusuf — Mon, 23 Jun 2025 23:50:52 +0000

Created for the purposes of entering the Google Cloud AI in Action 2025 Hackathon

tags: hackathon, ai, googlecloud, womenintech, adkhackathon

Introduction

The world is full of incredible women’s communities, but finding and connecting with them can feel impossible. For AI in Action 2025, we set out to solve this at scale by building MataConnect DataAgent, an intelligent, multi-agent system that automates the discovery, enrichment, and connection of women’s groups worldwide.

Why Data Agents?

Behind every powerful platform is high-quality data. MataConnect’s vision is only possible if we can continuously find, clean, and enrich information from a messy and fragmented web. That’s where our multi-agent approach comes in.

Building with the Agent Development Kit

I developed MataConnect DataAgent using the Agent Development Kit (ADK), which made it fast and modular to spin up autonomous agents, each with their own specialty:
Scraper Agents: Each agent targets a specific data source (like Google, Meetup, or Eventbrite) and pulls raw community information.
Enricher Agent: Takes the raw data, deduplicates entries, standardizes formats, and adds rich metadata such as location, category, and topics.

Tech stack highlights:

Gemini AI: For advanced enrichment and intelligent data matching.
MongoDB Atlas (hosted on Google Cloud): Secure, scalable storage for structured records.
Vertex AI Search: Powers real-time, intelligent search from the frontend.
FastAPI & React: For seamless API integration and a smooth user experience.

Why I Built It

Women shouldn’t have to dig through endless links to find their networks, and communities shouldn’t struggle to reach the people they can empower. MataConnect DataAgent bridges this gap by ensuring no community goes unseen.

Final Thoughts

I created this project specifically for the AI in Action 2025 Hackathon by Google Cloud, to showcase how modern agent frameworks and cloud technology can create real-world impact.

If you’re interested in building with agents, check out the Agent Development Kit docs and join the conversation!

adkhackathon

Thanks for reading! I would love to hear your feedback or connect with anyone working on similar challenges.

All about my Outreachy journey -- Wrap Up

Maryam Yusuf — Wed, 02 Mar 2022 20:59:33 +0000

I still find it hard to believe that my Internship is almost over. It is really true that time flies when you are doing what you really love.
The three past months have been a life changing experience for me.

In this blog post I would talk about my experience, my growth path and my future plan to keep contributing to the open source community.

What was one fear I had about the internship that turned out to not come true

In the contribution phase, I had about seven projects I was interested in because they require Python and I was conversant with it.

After going through each project on GitHub something stood out about the KCIDB project. It is core Python which I wanted, It is also centred around improving the Linux Kernel development lifecycle. I knew very little about linux system and a lot of things in the project so it was really scary deciding to work on the project. I made the decision to contribute knowing I will either succeed or fail. And if eventually I failed I will have lessons to reflect on.

Taking a task at a time instead of thinking of it as being something vast and complicated, helped me focus on things I needed at the moment to complete my task which made my journey really smooth

How did my mentor help me along the way?

I really appreciate the support and time that my mentor gave me in spite of his schedule, he took out time to answer my questions and guide me through my journey.
Thank you Nikolai for all the support
Thank you for making me see the end of this amazing program

An amazing thing happened

I am really glad that I got to work on another project within KernelCI. I worked on creating a pipeline service to send revision data to KCIDB. It was awesome as I got a deeper understanding of the KernelCI project.

Next Steps ?

I am so excited to continue this journey with my community. I look forward to building great stuff in the open source community.

I would love to offer any assistance to future interns and ready to provide any guide to the KernelCI project. Feel free to reach me if you have any questions about Outreachy.

It is never a goodbye to the Linux Kernel Community, I plan to keep contributing to improve the Linux Kernel development lifecycle. I cannot wait to see KernelCI achieve its aim of standardising hardware testing for the Linux kernel.

Career Interests

Maryam Yusuf — Mon, 07 Feb 2022 08:34:32 +0000

Hello everyone!

In this blog, I will be discussing my career interests and what I plan to pursue after the end of my internship.

What comes next after the internship

I believe I am writing this a little more enlightened and a little more ready for the next stage of my career. After working a while with the KernelCI project I fell more in love with the open source community and I would love to work on opensource projects at a professional level.

What I am looking for

Prior to starting my internship, I have worked as a Python developer for over three years. My last work involved building a SAAS(Software as A Service) application with Python and Django Framework. Then I decided to explore working in the open source community because that is where the future of software development lies 😊😊.

After my internship in March, I plan to look out for opportunities in the open source community.
I am willing to relocate as I am currently in Nigeria, and would love to also work remotely if I get the opportunity to do that.
I am looking for jobs mostly in Python, Site reliability engineering and Linux development, but I am willing to explore new areas, as this is what I love doing.

You can find me on Github or LinkedIn.

Modifying Expectations

Maryam Yusuf — Mon, 24 Jan 2022 09:07:52 +0000

Hi everyone!

In today's blog post I will be talking about all the progress I made over the past weeks, the hurdles I encountered and how my current progress has differed from what was planned originally.

Midway into my internship, I can gladly say that I have accomplished several milestones with my project. Over the past seven weeks there has been immense learning involved. Looking back at myself before the project I can see how much experience and improvement I have gathered.

Project Milestone

Back in November, when I applied for the KCIDB project I submitted a project timeline, to serve as a road map through my internship period. The timeline was reviewed by my mentor and he made some changes to make the process more realistic.

I spent a week familiarising myself with the project, to understand how KCIDB and its tools works and I closed a pull request I did not complete during my contribution phase.

For the remaining six weeks, I have been following the timeline with my mentor. Although we had to reprioritise the tasks and my mentor added some tasks for me to work on.

So far I have worked on :

Improving the CI/CD pipeline
Added test to the project
Updated the notification flow of the system
Learnt how Google cloud platform works, wrote test to monitor cloud functions, Pub/Sub topics, Schedulers, Firestore and BigQuery database deployment

Roadblocks

Some task took longer than expected. Working on the notification task was slightly challenging at first. It was necessary for me to understand how the project works, so as not to break any existing feature. I have discussed the struggles I faced in this blog-post.

These roadblocks delayed my progress. As I had to manually test my updates and was not sure what to check out for. Over this entire time, my mentor was very supportive of me.

The road further

The goal of the next half of my internship would be to add new features to KCIDB and write more tests to support the system.

I am immensely grateful to my mentor and the KernelCI community for being an integral part of my learning experience.

A look into my project: KernelCI

Maryam Yusuf — Fri, 07 Jan 2022 16:00:05 +0000

Hello everyone!

It has been five weeks now but it feels as if it all started a few days back only. 🙂

Today, I am going to explain my project i.e. KCIDB on which I am working on. This blog aims at helping newcomers to know more about KernelCI and how to get started.

What is KernelCI

KernelCI is a community-led test system focused on the upstream Linux kernel. It follows the open testing philosophy to enable the same collaboration to happen with testing as open source does to the code itself.

KernelCI aims to ensure the quality, stability and long-term maintenance of the Linux kernel.

To achieve its aim KernelCI has a lot of modules, which are broken down under different repositories

What I am building

One of the major component of the KernelCI is the KCIDB.

KCIDB is a package to collect all Linux kernel test results.
This includes tests run natively by KernelCI, Red Hat's CKI, Google's syzbot, ARM and many others.

One of the main tasks I’m working on, is to improve the notification flow. This enables submitters of test result to get their notification without failing, and enabling them set delayed notification for their submitted test data.

I’m also working on improving the CI/CD pipeline for the project, writing tests and providing support to make sure the flow for cloud deployment does not break.

How can you contribute to KernelCI?

KernelCI source code is designed to be publicly accessible.
If you want to start contributing to KernelCI you can easily find its project’s source code and issues on GitHub. Just choose a project of your interest and get started!!!

KernelCI projects on Github: https://github.com/kernelci

Everybody Struggles

Maryam Yusuf — Fri, 24 Dec 2021 15:42:52 +0000

Hi Everyone!

This is the second blog of my Outreachy series. It has been three weeks since my internship started at KernelCI and I am really enjoying the learning process.

This week I will be writing on my struggles so far and introduce an open source vocabulary term that I didn't know before.

So, coming down to the vocabulary term, let's see what a Github Action is !

GitHub Actions is a continuous integration and continuous delivery (CI/CD) platform that allows you to automate your build, test, and deployment pipeline. You can create workflows that build and test every pull request to your repository, or deploy merged pull requests to production.

Although I am not really a newcomer in the CI/CD world, but I have not written a workflow before. I have worked with projects that have workflows in them mainly on Bitbucket.
Writing my first workflow was thrilling, as I got to understand better how CI/CD works.

The Struggles 😭

My second and third week has been challenging. When I started contributing to KCIDB, I had very little knowledge of shell scripting, Google Cloud Platform and Kernel development. I had a lot to take in and figure out. Coming from a pure python and web background, it was a real struggle working on KCIDB project that required at least moderate knowledge of shell scripting and a good grasp of Kernel development.

My little safe place

I was initially very hesitant to ask for help because I'm afraid of asking stupid questions. But I realised that it is better to ask for help than to get stuck.

In conclusion...

I’m thankful to my mentor Nikolai who patiently explain things to me whenever I get stuck. He is super supportive and amazing and ready to answer all the questions I have no matter how pointless I think it might be.

No matter the challenges you face, Please do not give up!
Be adventurous in your journey of searching for knowledge and improving your skills.

I will leave you with this quote from Outreachy

It's important to remember that no one starts out being an "expert". Even your mentor was a newcomer to open source at one time. We all start out knowing nothing. We all learn, practice, and grow.
Learning can be hard! We may struggle to understand something. We may make a lot of mistakes when we practice new skills. And that's okay, because everyone struggles

Journey to Open Source with Outreachy

Maryam Yusuf — Fri, 10 Dec 2021 07:17:48 +0000

Hello! I'm Maryam Yusuf from Nigeria. I have been selected as one of the 61 interns of the Outreachy (December - March) '21 round.

I will be working on Improving the Unified Kernel Test Result Database System ("KCIDB") a package for submitting and querying Linux Kernel reports. KCIDB is a KernelCI project. The goal of KernelCI is to use an open testing philosophy to ensure the quality, stability and long-term maintenance of the Linux kernel.

As part of my 3-month internship, I'll be blogging to share my experiences with Outreachy and KernelCI. I hope this will help anyone who is interested in applying to Outreachy, or interested in tech.

What is Outreachy?

Outreachy is an initiative that provides paid remote internships with open-source software projects to the underrepresented groups in Tech. Their aim is to bring diversity and inclusivity to the tech industry.

My core values

My core values are adventure, community, and growth.

Adventure: I consider myself an adventurous person and value being adventurous. I love trying new things and taking risk on improving myself and helping the world to be a better place. I have a lot of exciting ideas piled up that I would try one day.
Community: Community engagement brings drastic change and impact to the world. Imagine a world without the the open source community, we would not have achieved what we have today.
Growth: I used to be so scared to step out of my comfort zone especially after rejection. I was scared to reapply for Outreachy when I got rejected in the last round, but I realise rejections are also part of growth.

Motivations for Applying to Outreachy

I came across Outreachy while I was exploring the Open source community after graduating from the University. I tried applying in 2020 but got rejected. One of my friends got the 2021 round for May, and encouraged me to apply again. I am definitely grateful to Hasana who encouraged me and helped out through my initial application process.

The second stage of the application was pretty easy as I am familiar with Python and the project I planned to contribute on (KCIDB) was also in Python. Although I had to learn a lot of things like the Google Cloud Platform(GCP) but it was interesting as learning new things is something I love.

I’m glad I didn’t let rejection stop me or else I wouldn’t be writing this blog post today.