Weaving social fabric at scale
Broadcast channels. You scroll through noise, adverts, and irrelevance. Finding something personally relevant is near zero.
Go quiet or get dominated by a few voices. Important requests get buried. The same 5 people do everything.
Complaint-driven. More neighbourhood watch than neighbourhood connection. 13M+ users but no matchmaking.
Nothing proactively connects people. If you want a running buddy, a book club, someone to share childcare, or a neighbour who can fix a boiler — you're on your own.
The UK government recognises loneliness as an epidemic. Post-pandemic, people live near each other but don't know each other. Community events rely on one overworked volunteer who eventually burns out.
The root cause: Coordination is expensive. A brilliant community organiser could solve this — but that person doesn't scale. Most communities don't have one at all.
Source: DCMS Community Life Survey / Campaign to End Loneliness
WhatsApp. Telegram. Chat. No new apps. No behaviour change.
Checks in through natural conversation. Builds a living profile — who you are, what you need, what you can offer. Not a static form.
Finds complementary people in your community. Explains why they're a good match. Brokers introductions with mutual opt-in.
Suggests meetups and activities based on real interests. Delegates logistics. "12 people want a book club — who can host?"
AI introduces itself. Checks in with each member via DM. Builds initial profiles.
"You and Sarah both run on Tuesdays and live near each other. Want me to connect you?"
Suggests first events from discovered interests. Book clubs, running groups, skill shares.
Playbooks emerge. Welcome Wagon for new neighbours. Skill Share for swapping expertise.
Reusable coordination patterns:
Welcome Wagon · Skill Share · Event Orchestrator · Interest Matcher
The need is intensifying and the technology just became possible.
Rising automation and unemployment mean society needs new ways to create meaning, connection, and economic activity beyond traditional jobs. Community is the foundation of whatever comes next.
Remote work + displacement from traditional employment = fewer incidental social encounters. The casual collisions that build trust and spark collaboration are disappearing.
Government-recognised, post-pandemic. 3.8M chronically lonely in the UK. Funders are actively seeking interventions.
People leaving Facebook, distrusting algorithms, but still wanting connection. There's a gap.
LLMs can genuinely understand what people mean from natural conversation — not just keywords. Matching quality is finally good enough to be useful.
Targeted, personalised messaging replaces broadcast. Instead of 50 people seeing irrelevant posts, each person gets what's relevant to them. Better signal-to-noise ratio than any group chat.
WhatsApp/Telegram penetration means zero app-download friction. Meet people where they already are.
Councils and grant bodies are actively funding community-strengthening tools. The money is looking for this solution.
The core technology works. What's needed is adapting it for the community use case and running a proper pilot.
A Telegram bot that builds profiles through conversation, matches people using vector similarity, and brokers introductions with explanations of why they'd get on. The matching engine is real and tested.
Next.js · Supabase · pgvector · Telegram Bot API
A 56-page product playbook covering onboarding flows, matching logic, community coordination patterns, and the full roadmap from single group to multi-community platform.
Landing site and algorithm test rig also built
| Product | Approach | Threat |
|---|---|---|
| Nextdoor | Complaint-driven, own platform | High if they add AI |
| Neya | AI community connection, own platform | Direct competitor |
| Lunchclub | AI networking events | Professional only |
| harmonica.chat | AI facilitation surveys | Different category |
Pricing is always per individual. But people belong to multiple groups — their neighbourhood, their running club, their workplace alumni, their council ward. One person, many contexts. Each group pays separately.
| Tier | Price | Who Pays | Example |
|---|---|---|---|
| Community | £1/mo | Per member | Villages, neighbourhoods, parishes |
| Organisation | £3–10/mo | Per member | Accelerator cohorts, conferences, alumni |
| Council-funded | Free to members | Council pays per community | Local authority resilience programmes |
One person in their local community (£1/mo), their workplace alumni group (£5/mo), and their council ward (free, council-funded) = three paying group memberships from a single user. Same profile, richer matching across contexts.
Like being in multiple Facebook groups — except each group is a paying customer.
Each new member makes matching better for everyone. Profile depth compounds across groups. Communities adopt collectively — switching means losing everyone's connections.
Abelique's true success isn't controlling behaviour — it's that the genuine human connections it catalyses persist even after the app itself fades.
— "Better Living Through Algorithms" by Naomi Kritzer
Hugo Award for Best Short Story, 2024 · Clarkesworld Magazine
Build something so good at connecting people that eventually the connections sustain themselves.
Four-time founder. Twenty years building coordination technology — from civic technology to digital democracy to AI-powered market intelligence.
Previously founded Represent.me — a digital democracy platform used in 61 countries with 1.2M+ votes cast. The Times called it "plotting a revolution in the way voters engage with politics."
We'll share milestones as Abelique develops.
[email protected] · London, UK