Watch the video @ 2x speed because our interactions take a while.

Inspiration

The thing that inspired us to create ArchieTech was having to repeatedly come up with the same structure for projects we work on. We were tired of mentally reconstructing similar architecture flows for apps that share common patterns. Rather than reinventing the wheel each time, we built an application to automate this process and guide developers through the entire project lifecycle.

What it does

ArchieTech will take you further than a to do list. ArchieTech is an intelligent agent that walks with you through every step of your programming journey. When beginning a project, you can have a natural conversation with Archie, our main AI agent. Provide as much or as little detail as you want about your project, and Archie will internalize and organize everything into a structured, actionable roadmap. Archie even suggests improvements and identifies quick wins to help you build momentum early.

How we built it

We built ArchieTech from the ground up, dividing responsibilities among our team members to focus on the most critical features while emphasizing scalability and user experience. Our tech stack leverages Google ADK for agent-to-agent (A2A) communication, enabling conversational interactions that feel natural and productive.

Challenges we ran into

Deployment complexities - Deployment is always challenging, especially within a 3-day hackathon timeframe. We encountered multiple hurdles including package management (pip) conflicts, DNS configuration issues, and Unix user permission problems that required creative problem-solving under pressure.

Designing cohesive A2A user interactions - Figuring out the conversation flow took several hours of iteration before we reached a design we were happy with. Even then, we had to revisit and refine our prompts multiple times to create an interaction that felt intuitive and helpful rather than mechanical.

Mastering Google ADK - Despite Google ADK's power as a development tool, it has a steep learning curve. Even though we've used it on two previous projects, we still discovered knowledge gaps that initially slowed our progress. Each project teaches us something new about its capabilities.

Accomplishments that we're proud of

We deployed an application in ~36 hours - from concept to live deployment, fully functional and ready for users.

We created a tool we actually want to use - This project solves a real pain point we experience regularly.

The potential for this project genuinely excites us - We see countless directions we could take ArchieTech, and we're eager to continue development.

What we learned

Prioritizing ruthlessly for deadlines - Not every feature can make it into v1, and learning to identify MVPs is crucial.

The power of effective teamwork - Clear communication and well-divided responsibilities allowed us to move fast without stepping on each other's toes.

Customer-first thinking - Since we are the target audience for this project, it was natural to build exactly what we needed, and we believe other developers will benefit just as much.

What's next for ArchieTech

We have ambitious plans for ArchieTech's future. Given the power and potential of this application, we're seriously considering monetizing it to support continued development and maintenance. Want to stay updated on our journey? Connect with us on LinkedIn, we'd love to share where this project takes us and get your feedback as we grow.

https://www.linkedin.com/in/arguellesalex/ https://www.linkedin.com/in/henrique-pitta-laranjinha-19594b249/ https://www.linkedin.com/in/bryanafernandez/ https://www.linkedin.com/in/george-ulloa-/

Built With

Share this project:

Updates