REMINAI

Event summarizer
Sign up page
Dashboard

Inspiration

Watching grandparents struggle to stay connected with their families inspired ReminAI. Modern families are scattered—kids in different cities, everyone buried in work. Grandparents want to know about their grandchildren's lives but navigating Instagram, Facebook, or group chats feels overwhelming. Meanwhile, busy parents and young adults feel guilty about not calling enough but lack the time for lengthy updates. We saw an opportunity: what if staying connected required minimal effort from everyone?

What it does

ReminAI bridges generational gaps through effortless storytelling. Family members snap photos and add brief notes about their day—a child's soccer game, lunch at a new restaurant, a work milestone. Our AI transforms these fragments into warm, readable narratives that grandparents can enjoy without technical barriers.

Elderly family members simply open ReminAI to see a Netflix-style interface showcasing recent family moments. Each person has their own profile with visual timelines of their activities. The AI-powered audio feature is crucial—using ElevenLabs voice synthesis, stories are narrated in natural, expressive voices. Grandparents with vision difficulties or who prefer listening can hear their family's updates read aloud with human-like intonation, making the experience accessible regardless of tech comfort level.

No complex navigation, no endless scrolling through confusing feeds, no missed updates buried in group chats. Just beautiful stories about the people they love, updated automatically, available to read or listen to at the press of a button.

How we built it

Frontend: React with TypeScript, deliberately designed with accessibility in mind. Large touch targets, high contrast text, intuitive navigation patterns borrowed from Netflix—an interface familiar even to less tech-savvy users.

Backend: Django REST Framework managing multi-profile authentication and family connections. Each family member journals independently; their stories appear in shared family timelines.

AI Integration:

Google Gemini 1.5 Flash analyzes photos and text to generate warm, conversational narratives. We crafted prompts specifically to create storytelling voices that feel personal and engaging—like hearing updates from a loved one, not reading auto-generated content.
ElevenLabs Text-to-Speech converts the AI-generated summaries into natural-sounding audio. The quality difference matters—robotic TTS would break immersion, but ElevenLabs' voices sound genuinely human, making elderly users feel like someone's actually talking to them about their family.

Accessibility Focus: One-click audio playback ensures grandparents can listen to updates hands-free, crucial for those with limited mobility or vision impairment.

Challenges we ran into

Balancing simplicity with functionality: We wanted rich features for younger users (easy photo uploads, quick note-taking) while keeping the viewing experience dead simple for elderly users. The solution was profile-based UX—content creators get powerful tools, content consumers get clean, accessible displays.

AI tone calibration: Early Gemini outputs felt robotic. Grandparents don't want clinical summaries; they want to feel connected. We iterated extensively on prompts to generate narratives that sound like a family member sharing their day, not a news report.

Audio integration complexity: Initially we used browser-native text-to-speech, but the quality was poor—monotone, mispronounced words, awkward pacing. Integrating ElevenLabs required building a separate audio pipeline but the quality improvement was dramatic. The voices sound natural enough that elderly users forget they're listening to AI.

Image loading performance: With multiple family members posting daily, the timeline needed to load smoothly even with many photos. We implemented lazy loading and thumbnail generation to maintain responsiveness.

Cross-generational design: Buttons and text that work for 25-year-olds can be illegible for 75-year-olds. We tested sizing, contrast ratios, and navigation patterns to ensure true accessibility.

Accomplishments that we're proud of

Creating an actually accessible family communication tool. Most apps claim accessibility but design for tech-savvy users. We built ReminAI for grandparents who've never used a smartphone app comfortably.

The ElevenLabs audio narration transforms the experience for elderly users. The natural voice quality creates genuine emotional connection—one team member's grandmother has macular degeneration, and being able to listen to family updates in a warm, human voice rather than struggle to read them would genuinely improve her quality of life.

Zero learning curve for consumers: Elderly users don't create content, they consume it. We eliminated every barrier—no account setup confusion, no complex menus, just "here are stories about your family" with a single play button.

Building a privacy-first multi-profile system where families control their own data, not a social media company mining it for ads.

What we learned

Accessibility isn't an afterthought: Designing for elderly users from day one made us better designers. Large buttons, clear typography, and intuitive navigation benefit everyone, not just seniors.

AI's true value is removing friction: Gemini doesn't just generate summaries—it eliminates the guilt barrier. Young people feel bad about not updating grandparents, so they don't. When updates require zero effort beyond living their lives, connection happens naturally.

Voice quality matters enormously: The jump from browser TTS to ElevenLabs was night and day. Elderly users engage more when the voice sounds genuinely human. Small technical details create massive UX differences.

Loneliness is a solvable technical problem: Social isolation among elderly populations is a health crisis. Technology created part of this problem (families scattered globally, communication moving to platforms seniors can't navigate). Technology can also solve it.

What's next for ReminAI

Simplified onboarding for elderly users: QR codes that family members can scan to instantly set up a viewing-only account for grandparents. Zero manual setup, zero confusion.

Smart digests: Weekly audio compilations using ElevenLabs—"Here's what happened in your family this week"—perfect for Sunday morning listening over coffee.

Voice personalization: Let families choose from different ElevenLabs voices, or even clone a family member's voice for narration.

Two-way communication: Let grandparents record voice messages back using speech-to-text. Keep it simple—one button, record, done. No typing required.

Health check-ins: Gentle prompts for elderly users to share how they're feeling. Family members get subtle alerts if grandma hasn't engaged with the app in several days.

Memory preservation: Auto-generate annual family albums combining everyone's stories, with audio versions grandparents can listen to like audiobooks.

Caregiver integration: Allow nurses or home health aides to help elderly users engage with the app, with proper permissions and privacy controls.

Translation: Multilingual support for immigrant families where grandparents speak different languages than grandchildren—ElevenLabs supports multiple languages.

The ultimate vision: eliminate loneliness in aging populations by making family connection effortless for everyone involved. Technology should bring families together, not drive them apart.