TechBuddy

Your patient, always-available AI companion helping seniors detect scams, master their devices, and understand AI — no app download required.

Comment

Inspiration

Millions of senior citizens are left behind by technology that moves too fast and help desks that feel too impatient. GenLink's mission hit close to home — seniors deserve a tool that meets them where they are: calm, respectful, and jargon-free. We saw an opportunity to build something that could realistically live in a senior center and actually get used every day.

What It Does

TechBuddy is a web app built specifically for senior citizens navigating modern technology. It offers five core features:

  • Scam Detector — Paste any suspicious text, email, or message and get an instant safe / scam / suspicious verdict in plain language, with clear next steps.
  • Help With My Phone — Step-by-step guided walkthroughs for everyday smartphone tasks like making video calls, taking photos, connecting to WiFi, and taking screenshots.
  • What Is AI? — A friendly, no-jargon chat interface that explains artificial intelligence concepts at a comfortable pace.
  • Practice Mode — Safe roleplay scenarios against common scam scripts (grandparent scam, tech support fraud, etc.) with an encouraging debrief afterward.
  • Quick Tips — One digestible tip per day on device safety, privacy, and AI literacy.

No installation, no login, no app store — just open a browser and go.

How We Built It

We built TechBuddy as a full-stack web application using Next.js 15 with TypeScript for both the frontend and API routes. The UI uses Tailwind CSS with large fonts, high contrast, and generous spacing — all deliberate choices for senior usability. AI features are powered by Z.ai's OpenAI-compatible API, allowing us to deliver real, context-aware responses for the scam detector, phone help walkthroughs, and AI explainer. The app is deployed on Vercel for a reliable, zero-config public URL that works on any tablet or desktop browser.

Challenges We Ran Into

  • Designing for accessibility first — It's harder than it sounds. We had to constantly fight the instinct to add features and instead ask "would an 80-year-old feel comfortable clicking this?" Every button, font size, and color choice was evaluated through that lens.
  • Making AI responses feel safe and calm — Generic LLM outputs can be verbose and alarming. We spent significant time on prompt engineering to ensure TechBuddy's tone stays reassuring even when flagging a scam.
  • Scam detection nuance — Real scam messages vary wildly. Building a prompt that produces accurate, consistent verdicts without over-alarming users on benign messages required careful iteration.

What We Learned

Building for a non-technical audience forces a kind of discipline that makes you a better developer. We learned that good UX for seniors is really just good UX — clear language, obvious actions, and no dead ends. We also got hands-on experience integrating OpenAI-compatible APIs within the Next.js App Router architecture.

What's Next

We'd love to add voice input so seniors can speak their questions instead of typing, a caregiver dashboard so family members can see what topics their loved ones are exploring, and a multilingual mode to serve non-English-speaking senior communities.

Built With

Share this project:

Updates