IRIS | Devpost

Introduction
4 Key Solution
Suppported SDG's
Behind the Band
Our Flowchart
Closing

Inspiration

Our inspiration stems from the myriad of challenges faced by women, both inside and outside their homes. Safety should be a fundamental right for everyone. With the advancement of AI, we saw an opportunity to address traditional issues that cause harm to victims, such as providing psychological support, preserving evidence, and reporting to authorities.

What it does

Meet IRIS, an innovative mobile application designed to provide comprehensive security and user convenience, utilizing the latest advancements in AI Solution. With a focus on preventing and responding to domestic violence and sexual assault, IRIS offers a holistic approach to safety and evidence collection. The IRIS app, when paired with a smart band, detects emergencies through abnormal motion detection, voice sentiment analysis, and heart rate monitoring. It provides immediate assistance, records crucial evidence, and offers a platform for psychological support and incident reporting.

How we built it

We developed a mobile application using the Next.js framework with Capacitor to ensure cross-platform compatibility on the front end. For the back-end architecture, we utilized Go, GORM as the Object-Relational Mapping tool, Go Fiber as the web framework, and Supabase as our databaser. We chose Go because it offers significantly faster processing times compared to other languages. After that, we also integrated OpenAI's API as the base large language model (LLM) and employed frameworks such as Langchain, LangGraph, and LangSmith. These integrations enable a robust LLM chatbot system with multi-agent capabilities, functioning as a document writer, personal psychological consultant, and an agent for integration with other applications that suitable with our purposes.

Challenges we ran into

We faced several challenges, including time constraints and the complexity of integrating IoT devices with the mobile application. Ensuring seamless communication between the app and the smart band required extensive troubleshooting and optimization.

Accomplishments that we're proud of

Despite the challenges, we successfully createdvv a system that will significantly help many people by leveraging AI, IoT, and mobile app integration. Our mobile app functions as intended and is built on a solid architecture. Additionally, we fine-tuned the pretrained LLM from OpenAI's API using prompt instructions to ensure each agent performs its specific role effectively. It was quite satisfying to see the application provide individualized support and potential employment opportunities.

What we learned

Through this project, we learned how to create a functional system based on a real-world issue that is deeply personal to us. We gained valuable insights into leveraging cutting-edge technology, including AI and IoT, to innovate and address previously unmet needs in the realm of personal safety and emergency response. We discovered the importance of integrating various technological components, such as large language models and smart sensors, to build a cohesive and efficient solution.

Most importantly, we learned that by aligning our efforts with global missions, such as the United Nations Sustainable Development Goals (SDGs), we could make a meaningful impact. This alignment helped us stay focused on creating a solution that not only addresses immediate safety concerns but also contributes to broader objectives like gender equality, good health and well-being, and the promotion of peaceful and inclusive societies. Our experience underscored the significance of combining technological innovation with a mission-driven approach to create sustainable and impactful solutions.

What's next for IRIS

Moving forward, we will diligently pursue our goal by continuously enhancing app features and optimizing the user experience, particularly in AI development and IoT improvement. We plan to introduce advanced predictive analytics to anticipate potential threats based on user behavior and environmental data, further increasing the proactive capabilities of IRIS. To foster a community-driven approach to safety, we will implement a feature that allows users to share safety tips and report hazardous areas in real-time, creating a crowdsourced safety map accessible to all users. This map will provide valuable insights and help users avoid dangerous locations. Additionally, we aim to collaborate closely with authorities and local organizations to create a safer world for everyone. We will establish partnerships with law enforcement and emergency services to ensure rapid response times and seamless integration with our system. By developing a network of trusted partners, we can provide users with immediate and effective support during emergencies. To amplify our impact, we also plan to launch educational campaigns and workshops focused on personal safety and the responsible use of technology. These initiatives will empower communities with the knowledge and tools they need to stay safe and support each other.