Inspiration
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for AlnonAI
AlnonAI: Making Support Easier to Reach
Inspiration
Alcohol-related harm rarely affects just one person. It can affect families, relationships, financial stability, education and access to healthcare. Yet finding the right kind of support can be difficult, especially when people are unsure where to begin or feel uncomfortable asking for help.
That was the problem behind AlnonAI. I wanted to build something that could help people navigate difficult circumstances without making them explain their entire situation to multiple organisations just to find basic assistance.
I also wanted to approach the problem with privacy and accessibility in mind. Seeking help should not have to become another source of fear or embarrassment.
What it does
AlnonAI is a discreet, AI-powered healthcare and support navigation platform designed for individuals and families affected by alcohol-related harm.
It brings several forms of assistance together in one place:
- AI Support Assistant: Helps users find relevant information and understand possible next steps.
- Healthcare Navigation: Makes health information easier to understand and helps users prepare for conversations with healthcare professionals.
- Emergency and Safety Support: Provides emergency resources, trusted-contact tools and safety guidance.
- Support Directory: Helps users discover relevant NGOs, rehabilitation services and other support organisations.
- Legal and Financial Guidance: Connects users with information about legal aid, government schemes and financial assistance.
- Education and Employment Support: Helps users explore opportunities to rebuild stability and independence.
- Accessibility: Includes language and interface options intended to make the platform easier to use.
AlnonAI is designed to complement professional healthcare and social services, not replace them.
How I built it
I began with the original AlnonAI concept and expanded it into a healthcare-focused version.
The current prototype uses Python and Streamlit for the interface, with SQLite for local data storage. Its support assistant uses a retrieval-based approach to find relevant information from a small knowledge base. The application also includes resource discovery, contact management and safety-related workflows.
I focused on making the different sections useful as actual workflows rather than simply presenting information on a dashboard.
I also considered how sensitive information should be handled. The prototype keeps conversations within the current session rather than treating private conversations as ordinary application data.
Challenges I faced
One of the biggest challenges was balancing accessibility, privacy and usefulness.
A platform dealing with sensitive family circumstances cannot assume that every user has the same level of privacy or feels comfortable openly seeking assistance. This influenced the discreet design and the emphasis on user-controlled safety features.
Another challenge was making the healthcare component responsible. An AI assistant should not diagnose conditions, prescribe medication or give users a false sense of medical certainty. AlnonAI therefore focuses on general health information, understandable explanations and pathways towards professional assistance.
I also faced technical challenges while setting up the application locally, particularly Python environment compatibility and package installation on Windows. Working through these issues helped me understand dependency management, virtual environments and application deployment more practically.
What I learned
Building AlnonAI taught me that developing technology for social impact involves much more than writing code. It requires understanding the people who might use it, considering the risks of the technology and designing around real-world limitations.
I learned to think about privacy as part of the user experience, healthcare information as something that needs careful sourcing and AI as a tool that should support human decision-making rather than replace it.
Most importantly, I learned that a useful solution does not always need to solve every part of a person's problem. Sometimes, helping someone find the right first step is meaningful in itself.
What's next
AlnonAI is an evolving prototype. My next priorities are expanding its verified healthcare and support knowledge base, improving multilingual assistance, strengthening data protection, testing accessibility and validating its safety workflows with appropriate professionals.
My goal is to develop AlnonAI into a reliable bridge between people who need support and the organisations equipped to provide it.
AlnonAI is built around one simple idea: finding help should be easier than living without it.
