At SCaLE 22x, the impact of AI and smart infrastructure was discussed and demonstrated. In just a few weeks, teams built what would normally take months or years of discussions, workshops, roadmaps, user groups, and complex procurements. With the guidance of mentors and industry leaders, hackers turned ideas into working solutions, proving that open-source, decentralized AI can solve real-world problems much faster and more effectively than with traditional systems and procurement methods.
The Role of Mentors: A Follow-Up to “The Hackers & Mentors Redefining Smart Infrastructure”
A previous post explored the importance of mentorship in guiding AI innovation. This hackathon provided a real-world case study of those principles in action, where experienced mentors helped shape projects into viable, impactful solutions. Rather than blindly building apps and using AI for AI’s sake, this was a deliberate, agile, yet methodical demonstration of how the world should operate today—focusing on people and outcomes.
Guiding the Future: Insights from Jon Connors
Jon Connors, an advisor for KWAAI and leader of the hackathon, framed the challenge as follows: “We already have the data—Ring cameras, sensors, smart buildings—but they don’t talk to each other. Imagine if they worked together to guide evacuations.” The winning teams took this vision and built decentralized, AI-driven tools for emergency response.
- Watch Jon Connor’s Message for the Industry and about how the Hackathon was organized
Meet the Winning Teams
These teams didn’t just pitch ideas—they built functional, AI-powered solutions that have the potential to redefine disaster response. Some teams were even formed on the fly, as mentors like Chuck Chekuri helped bring together individuals with the right skills to collaborate in real-time.
1. DIY Parties – Community-Driven Wildfire Resilience
Problem: Individual homeowners struggle to prepare for wildfires, and community engagement is low.
Solution: DIY Parties gamifies community wildfire preparedness by organizing local events supported by AI-generated risk assessments, automated event planning, and sponsorship tracking.
Yeled Omercy, the sole coder behind DIY Parties, built a fully functional platform in just two weeks—something that would have taken months or years without AI.
“I had no front-end web experience, but AI-assisted coding enabled me to rapidly prototype, debug, and automate—letting me focus on the problem, not the tools.”
Team:
- Yeled Omercy, CEO HOPEHUB Building DIY PARTIES
- Jason Barto – Fire preparedness expert, provided validation and community engagement strategies.
- Brad Knowles – Firewise committee member, advising on early adoption.
- Ginie Servant-Miklos, Assistant Professor Climate Psychology
- Erik Cortes – CEO Sherpas, Biz Advisory Consultant
Key Features:
- AI-generated event templates tailored to wildfire risks.
- Sponsorship tracking to fund resilience efforts.
- Community-driven approach to increase participation.
Inspired by time in Zambia, Yeled Omercy saw how strong community bonds helped people navigate crises—something often missing in the U.S. DIY Parties aims to change that by making wildfire preparedness a shared, action-driven experience.
As the sole coder, Yeled built the platform using AI to accelerate development, securing first place at the SCaLE hackathon. Despite having little front-end web experience, AI-assisted coding enabled rapid prototyping, debugging, and automation—allowing a working platform to emerge in just two weeks.
This represents a major shift in the industry, where AI is no longer just an optimization tool but a fundamental enabler, reducing barriers and empowering individuals to build real-world solutions at unprecedented speeds.
The traditional model of development—where teams of specialists are required to bring an idea to life—is rapidly changing. Now, a single person with vision, determination, and AI-assisted coding can achieve in weeks what used to take months or years.
This is not a future possibility—it is happening right now. Developers, architects, engineers, and industry leaders need to recognize that AI is no longer just an enhancement; it is a disruptive force that is rewriting the rules of how solutions are built. Those who ignore this shift risk being left behind. Those who embrace it will unlock new levels of speed, efficiency, and creativity.
AI is also embedded in the DIY Parties app itself, generating event templates based on location-specific risks, supporting community agreements, and streamlining the planning process. Instead of replacing experts, the platform connects people with the right knowledge and resources—demonstrating that disaster resilience can be built collaboratively, efficiently, and even enjoyably.
The success of DIY Parties is proof that AI is no longer an experimental tool—it is the tool. This is not just about software development; it’s about how we solve problems at scale. The industry must understand, accept, and integrate this reality now—or risk being outpaced by those who do. The future of development isn’t coming—it’s already here. The only question is: Who’s ready to build in this new reality?
Mentor:
- Matt Small – Go-to-market mentor, helped refine focus on the preparedness gap.
- Watch the Short Interview with Yeled about DIY Parties
- Watch the Full Interview with Yeled about DIY Parties and what inspired the development
- Sign up for beta and provide input here to get updates on DIY Parties
2. AidPulse – AI-Driven Emergency Response Network
Problem: During disasters, responders often lack real-time data on what victims actually need, leading to inefficient resource distribution.
Solution: AidPulse is an AI-powered, peer-to-peer emergency response system that operates on edge devices, ensuring functionality even when networks are down.
Team:
- Aakarsh Surendra – Founding Product Manager, helped assemble the group.
- Manan Shah – Lead developer, focused on AI-driven analytics.
- Sachin Lodhi – Head of AI, Key contributor to development and system design.
- Jayde Mitchell – Product Engineer
- Craig Albuquerque – Lead Analytics and AI Developer, Assisted in refining the platform’s technical capabilities.
Key Features:
- AI-driven contextual awareness to match responders with real needs.
- Privacy-first, decentralized communication.
- Works offline when traditional networks fail.
The AidPulse team came together to develop a solution that ensures critical information reaches the right people at the right time. Using open-source tools, the team prioritized accessibility and adaptability, believing that disaster relief technology should be both efficient and widely available.
A major challenge in disaster response is the misalignment between what victims need and what aid organizations provide. AidPulse addresses this by using AI-driven analytics trained to be contextually aware. The system prioritizes real-time needs, allowing first responders and policymakers to allocate resources efficiently. Importantly, AidPulse works even when communication networks fail—leveraging edge computing to ensure seamless information flow in critical moments.
Guided by an AI expert mentor with experience at HUD, the team refined their approach throughout the hackathon, ensuring the solution was both practical and scalable. The result was a fully functional prototype that earned second place, proving the potential of AI-driven, decentralized emergency response.
Mentor: Chuck Chekuri
Chuck guided the AidPulse team through the hackathon and helped bring members together at the start. His ability to quickly form a team and ensure they focused on real-world impact was key to their success. “Building apps is cheaper and faster than ever before. The real challenge is mentorship and ensuring teams focus on real-world impact.”
- Watch Chuck Chekuri in an interview about how he mentored AidPulse
- Watch Manan Shah in an interview on how the team created AidPulse
3. Deeper Breath – Smart Wearables for First Responders
Problem: First responders need real-time air quality data to navigate hazardous environments safely.
Solution: The Canary Mask is an AI-powered smart wearable that detects air quality threats and provides real-time data to responders.
Team:
- Dustin Wish – Founder, Global Executive in Technology and Media, Lead developer, specializing in consumer electronics and wearables.
- Fred Schechter – Co-founder, Industrial designer with expertise in air purification.
- Nanoz – Sensor Specialist – Provided expertise in advanced air quality detection.
Key Features:
- AI-powered sensors for biohazards, gases, humidity, and air pressure.
- Multiple connectivity options – Wi-Fi, Bluetooth, 5G, and LoRa.
- Health monitoring with oxygen and carbon monoxide detection.
- Smart filters (N99 & P100) with real-time monitoring for replacements.
Deeper Breath was founded by a team with deep expertise in consumer electronics, air purification, and wearables. The concept evolved from prior work in air filtration, inspired by the need for real-time environmental awareness during disasters. The team rapidly iterated on the Canary Mask, integrating edge AI to process data directly on the device, ensuring functionality even in network outages.
With a vision for broader applications, Deeper Breath has secured interest from military partners, Cal Fire, Fresno State University, and the Raiders football team, positioning the product for both emergency response and occupational safety. The project was further refined through mentorship from Victor H. Esch, a seasoned fire chief and disaster management expert, who guided the team in focusing on critical gaps in firefighter and emergency worker safety.
Mentor: Victor H. Esch – Victor, an experienced fire chief and emergency response expert, helped the Deeper Breath team refine their wearable mask concept, ensuring it addressed gaps in existing industry solutions.
- Watch Dustin Wish in an interview about Deeper Breath
- Watch Victor H. Esch on what advice he provided as a mentor
- Deeper Breath Web Site
- View the Automated Building Slide Commentary
Other Entries & Insights from Liam Broza
Tapestry – AI-Powered Digital Identity for Disaster Recovery
Problem: After disasters, many victims lose crucial documents—proof of identity, homeownership, and records needed for aid. Without them, accessing FEMA assistance, insurance, or legal claims becomes nearly impossible.
Solution: Tapestry Credential uses AI to reconstruct lost identities by compiling digital evidence from sources like phone records, satellite images, and transaction history, helping disaster victims verify who they are and what they own.
Team: Liam Broza
Key Features:
- AI-driven data reconstruction to replace lost documents.
- Uses phone records, social media, and public data to verify residency.
- Provides a certified credential to access aid and legal services.
Insights from the Tapestry Team:
I first saw Liam Broza present live during the Kwaai session, where he explained how Tapestry leverages AI-driven forensic data recovery to restore lost identity records. The next day, I was able to find Liam in person, and he was available for an in-depth interview.
Originally conceived after Hurricane Maria left thousands without proof of homeownership, Tapestry Credential was revived at the hackathon to harness AI for disaster recovery. By leveraging personal digital footprints, the system helps victims rebuild their records and access critical aid. The team envisions pilot programs and partnerships to bring AI-driven identity recovery to real-world disaster response efforts.
- Watch the interview with Liam Broza about Tapestry
- https://liambroza.com/
- https://www.companionintelligence.com/
Hackathon Videos:
Watch the rest fo the videos.
- Deeper Breath – Bronze
- DIY Parties – Gold
- Tapestry Credentials
- Firemeister
- AidPulse – Silver
- FireRiskAI
- Responder LLM
- Medusa A
- GoSafe
Looking Ahead: Smarter, More Resilient Cities
The SCaLE Hackathon was more than just an experiment—it was a wake-up call. The industry must acknowledge that AI-driven development is no longer theoretical; it is actively reshaping how we build solutions, who can build them, and at what speed. Ignoring this shift is no longer an option.
The results of this hackathon reaffirm the importance of open-source, decentralized AI in emergency response. The success of these teams also highlights the power of mentorship, which turns abstract concepts into working solutions.
Shaping the Future: Procurement and Digital Twins
The SCaLE hackathon proved that open-source, decentralized AI can drive real-world impact faster than traditional models. It also highlighted a critical challenge—our approach to procurement and development must evolve to keep pace with innovation.
This is just the beginning. The next post will explore how mentors, technologists, and industry leaders are breaking outdated cycles and reshaping the future of smart infrastructure, digital twins, and emergency response. Stay tuned.Join the discussion: How can buildings and AI work together to improve disaster response? Let’s keep pushing the boundaries of open innovation.
Read Part 1 about the Mentors that shaped the hackathon here.
