Inspiration
The loneliness epidemic is no longer a matter of anecdote, it is a clinically documented public health crisis. According to a December 2025 AARP study of over 3,200 respondents, 40% of U.S. adults aged 45 and older now report experiencing loneliness, a statistically significant increase from 35% in both 2010 and 2018. A Nature-published meta-analysis encompassing 126 studies and 1.25 million older adults established a global prevalence rate of 27.6%. The health consequences are severe: the U.S. Surgeon General's 2023 advisory equated the mortality impact of chronic social disconnection to smoking 15 cigarettes daily, while the National Institute on Aging has linked social isolation to a 50% elevated risk of dementia.
Critically, this crisis is not confined to older populations. The Cigna Group's 2025 national survey found that 57% of Americans across all age groups report loneliness, with Gen Z exhibiting the highest rates — 73% report feeling lonely sometimes or always — despite being the most digitally connected generation in history. Young people increasingly lack access to mentorship, intergenerational perspective, and the practical life skills traditionally transmitted through extended family and community networks.
These two demographics represent complementary halves of the same structural failure. Older adults possess accumulated wisdom, professional expertise, cultural knowledge, and lived experience, yet face shrinking social circles with each passing year. Young people are actively seeking guidance, skills, and authentic human connection, yet remain isolated within algorithmic echo chambers. The infrastructure to bridge this gap does not currently exist.
BridgeGen was conceived to address this precise deficit: a technology-mediated platform that facilitates genuine intergenerational relationships, grounded in the peer-reviewed evidence base demonstrating the efficacy of structured intergenerational programming.
What it does
BridgeGen is a web-based platform that leverages artificial intelligence to facilitate reciprocal skill exchange and mentorship between socially isolated older adults and young people. The platform employs three integrated technical systems designed to lower the barriers to intergenerational connection while preserving the authenticity of the resulting relationships.
Intelligent Compatibility Matching — The matching engine moves beyond surface-level attribute pairing. User profile narratives are processed through a sentence-transformer model (all-MiniLM-L6-v2) to generate 384-dimensional semantic embeddings. Cosine similarity scoring identifies candidate matches with genuine depth of overlap across interests, personality indicators, and skill complementarity. A constraint satisfaction layer then filters for logistical feasibility: schedule alignment, language compatibility, and geographic proximity. The result is a match optimized not merely for demographic similarity, but for the conditions under which meaningful human bonds are most likely to form.
Adaptive Communication Bridge — One of the most significant barriers to intergenerational technology platforms is the digital literacy gap itself. BridgeGen addresses this through a dynamic interface personalization system that observes user interaction patterns, click velocity, time-on-screen, navigation error frequency, and progressively adjusts the UI without requiring manual configuration. For users exhibiting lower digital fluency, the platform shifts to voice-first interaction, enlarged interface elements, and simplified navigation hierarchies. For digitally native users, the platform exposes richer functionality including calendar integration, notification management, and multimedia messaging. Both populations experience a platform that appears purpose-built for their needs.
Conversation Facilitation Agent — Research consistently identifies the initial interaction as the highest-friction point in intergenerational programming. BridgeGen's facilitation agent draws on both participants' profile data to generate structured, mutually beneficial activity prompts — for example, a skill exchange session where one participant teaches a family recipe while the other assists with video-calling setup. This scaffolding is designed to be temporary: after three to five successful interactions, the agent systematically reduces its involvement, allowing the relationship to transition to self-sustaining organic connection. This approach is directly informed by the intergenerational programming literature, which consistently demonstrates that structured early interactions with meaningful roles for both parties produce the strongest long-term relational outcomes.
Accomplishments that we're proud of
Validated a genuine market whitespace. Through systematic competitive analysis spanning Product Hunt, Y Combinator cohorts, and the broader eldercare technology landscape, we confirmed that no existing platform combines AI-powered intergenerational matching, adaptive accessibility interfaces, and structured conversation facilitation. Adjacent solutions either replace human contact with artificial companions (ElliQ, Replika), rely on manual coordination that cannot scale (DOROT), offer reverse mentoring without intelligent matching (Cyber Seniors), or operate as transactional companion marketplaces (Mon Ami). BridgeGen occupies a category of one.
Achieved triple UN Sustainable Development Goal alignment. BridgeGen directly addresses SDG 3 (Good Health and Well-being) by combating the documented health consequences of social isolation, SDG 10 (Reduced Inequalities) by promoting social inclusion irrespective of age in accordance with Target 10.2, and SDG 11 (Sustainable Communities) by rebuilding intergenerational social infrastructure eroded by urbanization and digital fragmentation.
Grounded every design decision in peer-reviewed evidence. A 2025 systematic review spanning eleven years of intergenerational program data demonstrated consistent improvements in older adult well-being, cognitive engagement, and social functioning. A 2026 scoping review encompassing 69 studies confirmed that programs incorporating structured interactions with meaningful participant roles produce the strongest measurable outcomes. BridgeGen's matching, facilitation, and fade-out systems are direct operationalizations of these evidence-based principles.
Conducted substantive pre-launch validation. Over 15 structured interviews with older adults at community centers informed both problem validation and interface design. Pilot pairs were recruited and tested, generating qualitative feedback and satisfaction data. A landing page accumulated 50+ organic signups, and a letter of interest was secured from a community center prepared to host a formal pilot program.
What we learned
The technology must be invisible to succeed. The most important insight from user testing was that older adults do not want to interact with "AI" — they want to be connected with a person who feels like a natural fit. Every technical decision, from the matching algorithm to the adaptive interface, must serve the human outcome without drawing attention to itself. The moment a user is conscious of the technology mediating their experience, trust erodes.
Older adults are a profoundly heterogeneous population. A 65-year-old retired software engineer and an 85-year-old widow living alone have fundamentally different technological capabilities, social motivations, and communication preferences. Our adaptive interface exists because early prototypes that treated "seniors" as a uniform category failed immediately in testing. Personalization at the individual level, not the demographic level, proved essential.
Institutional trust is a prerequisite for adoption. Community center interviews revealed that older adults are unlikely to adopt a platform encountered through digital advertising or app store discovery. Partnership with trusted local institutions, senior centers, libraries, faith communities, healthcare providers, is not merely a distribution strategy but a fundamental requirement for user acquisition in this demographic.
Facilitation timing determines relationship survival. The conversation facilitation agent's withdrawal strategy required significant iteration. Premature withdrawal resulted in relationship attrition as conversations lost structure. Delayed withdrawal produced a sense of surveillance that undermined relational authenticity. The optimal approach proved to be a graduated reduction calibrated to interaction frequency, message sentiment, and explicit participant feedback, a balance between scaffolding and autonomy.
Validated traction outweighs technical sophistication in impact-oriented competitions. Five authentic testimonials from pilot participants proved more compelling to early evaluators than detailed explanations of our NLP pipeline. This reinforced a broader principle: in social impact ventures, the proof is in the human outcome, not the architecture.
What's next for BridgeGen
Months 1–3: Expand to 50 active intergenerational pairs across three community centers in the Greater Atlanta metropolitan area. Apply to Georgia Tech's CREATE-X accelerator program for structured venture development support.
Months 4–6: Scale to ten institutional partnerships across metro Atlanta, encompassing community centers, senior living facilities, and municipal aging services programs. Pursue federal grant funding through the Administration for Community Living.
Months 7–12: Expand operations to five major U.S. metropolitan areas: New York, Chicago, Houston, and Los Angeles, through partnerships with Area Agencies on Aging. Launch premium subscription tier offering priority matching, video session recording for intergenerational memory preservation, and family sharing dashboards.
Long-term vision: Establish BridgeGen as critical social infrastructure: a platform where intergenerational connection is not contingent on geographic accident or family structure, but is accessible, intentional, and sustained by technology that brings people together and then deliberately removes itself from the equation.
Log in or sign up for Devpost to join the conversation.