Inspiration

The idea for Reconnect Generations came from a simple but painful truth: most families realise too late that they never truly preserved a grandparent’s voice, stories, or personality. We saw how people rely on scattered phone videos, short voice notes, or vague memories—none of which capture the way a grandparent spoke, laughed, told stories, or gave advice.

This loss becomes permanent, and future generations miss out entirely on their family’s identity, wisdom, and history. That emotional gap is what inspired us to build a tool that preserves a grandparent’s presence before it’s too late.

What it does

Reconnect Generations is a digital platform designed to preserve a grandparent’s voice, stories, and personality—and transform them into an interactive Memory Twin the family can talk to anytime.

It works by guiding grandparents through simple, AI-led interviews. The system records their real voice, models their communication style and values, and organises their stories into a secure memory archive. Families can then:

Hear answers in the grandparent’s actual cloned voice Revisit memories and stories Ask questions and interact naturally Access a preserved legacy for future generations

It’s an emotionally powerful, easy-to-use way to make sure a grandparent’s voice is never lost.

How we built it

We combined modern AI tools, cloud infrastructure, and simple UX principles to create a smooth experience for both grandparents and families.

Key tools and technologies used:

Hostinger ElevenLabs Google Cloud Platform Google Text-to-Speech API Google Gemini 2.2 Flash Lovable.ai V0 Vercel GitHub Stripe Meta Ads (Facebook Business Suite) ChatGPT-5 Nanobanana VEO3

We used AI for voice capture, transcription, personality modeling, and conversational responses. Everything else-hosting, deployment, payment processing, and frontend flow - was designed to be lightweight, secure, and scalable.

Challenges we ran into

Making the experience simple enough for elderly users Ensuring emotional safety around voice cloning and memory preservation Training voice models with limited initial recordings Managing latency and quality in the interactive Memory Twin Balancing deep AI features with a clean, non-technical interface Convincing early testers to share sensitive family memories Crafting messaging that respects the emotional weight of the product

These challenges forced us to focus on clarity, privacy, and trust.

Accomplishments that we're proud of

Built a full working prototype capable of real interactive Memory Twin conversations Successfully guided grandparents through AI interviews with minimal friction Achieved convincing voice responses using limited recordings Created a clean UI that works even for non-technical seniors Formed a compelling business model with early customer validation Designed a product that genuinely helps families preserve something irreplaceable

Most importantly, we watched families react emotionally when hearing a preserved voice or story. That validated everything.

What we learned

Preservation is not a “nice-to-have”; it’s a deeply emotional need Families hesitate because they fear it may be “too late” - urgency matters Seniors respond better to guided, conversational AI than manual recording tools Voice, personality, and story together create a far more meaningful legacy than simple audio clips Building trust is as important as building features

What's next for Reconnect Generations

Building a richer memory archive with photos, timelines, and generational mapping Adding multilingual European support Growing through partnerships with elder care homes, hospices. Improving voice accuracy with long-form recordings Creating a family “legacy vault” for secure long-term storage

Ultimately, our vision is to become the standard way World-wide families preserve and interact with the voices, wisdom, and stories of the people who shaped them.

Built With

Share this project:

Updates