Golden Builders - Global AI Hackathon: Uniting Innovators Worldwide
AI Tinkerers - Paris
Hackathon Showcase

Golden Builders

Team consisting of Ty (BTHA SYSTEM founder — Full‑Stack Web3: Solidity, Solana, XRPL, React/React Native) and Jennifer Delalande (16‑year AI engineer, EBW3NT, AI R&D, JavaScript).

2 members Watch Demo

https://aura-companion-tztjqd6.gamma.site/
AURA COMPANION is a gamified, evolving AI companion—a human Jarvis and a smart Tamagotchi—that drives users to take action in the real world: it helps them achieve personal and professional goals, develop their well-being, build genuine social and romantic connections, and make financial and educational progress. Its uniqueness is that it combines a life-wide dashboard with proactive voice/text coaching, meaningful gamification, and real-world actions—so progress only counts when users actually do things—creating a living mirror that grows alongside its user. Powered by a secure Jarvis-style agent, AURA plans, summarizes, and schedules with explicit consent, while Proof-of-Action (GPS, on-device photo, verified interactions) anchors results in reality; the avatar visibly evolves as follow-through increases. For judges, this delivers clear problem/solution fit (from screen-time to real action), measurable impact (POA/week, streaks, changes in pillar scores, IRL participation), and a reproducible, demo-ready architecture (React JS/Next.js → NestJS → LLM + STT/TTS → Supabase) presented with a concise, smooth live walkthrough.

Here’s the clean English version, matching your structure and tone:


AURA COMPANION — Pitch Deck (Hackathon)

1) Cover

AURA COMPANION
Less screen, more life: the AI that turns intentions into actions.
Jarvis + real-life Tamagotchi — a gamified, multimodal AI companion focused on real-world actions.


2) One-liner & Vision

One-liner: The first AI companion that rewards only real actions (Proof-of-Action) and plans for you via a secure Jarvis agent.
Vision: Build the life cockpit — a companion that measures, motivates, and materializes goals (well-being, work, learning, relationships, finances) and co-evolves with the user.


3) Problem

  • Social isolation, procrastination, decision fatigue.
  • Fragmented tools (well-being, productivity, finance, dating) → no single overview.
  • Existing AI coaches advise but don’t anchor real habits.

4) Why Now

  • Maturity of multimodal LLMs + sensors (GPS, health, camera).
  • Push for digital hygiene: less scrolling, more real life.
  • Need for a personal co-pilot that plans, motivates, and verifies IRL action.

5) Solution (Overview)

  • Gamified AI companion (evolving avatar).
  • Life cockpit: pillar scores + % of goals achieved.
  • Real-world quests and Proof-of-Action (POA).
  • Jarvis agent: plans, summarizes, organizes, connects (calendar, email, health, finance, events).

6) How It Works (4 steps)

  1. Onboarding: pillar goals, AURA’s voice/persona, coaching preferences.
  2. IRL quests (daily/weekly): micro-actions & social/love challenges.
  3. Proof-of-Action (POA): GPS, on-device photo, interaction (call/message), partner QR/NFC.
  4. Dashboard: 0–100 pillar scores, % goals, streaks, real/virtual ratio (+ voice/text feedback).

7) Key Features

  • Personalization: 2D/3D avatar, coaching style, voice.
  • Multimodal: chat + voice (STT/TTS), visual recaps.
  • Useful gamification: XP, levels, badges, power-ups, group challenges.
  • POA: GPS, local photo, social interactions, QR/NFC.
  • Autonomous Jarvis: planning, briefs, summaries, event suggestions, calendar add (granular consents).
  • Security/ethics: GDPR-by-design, safe-meet, anti-addiction.

8) Gamification “real > virtual”

  • XP earned only via POA.
  • Streaks, badges, and social power-ups.
  • AURA avatar blossoms with real progress.
  • Group challenges (city/campus) to spark connection.

9) Proof-of-Action (POA)

  • Validation: GPS (presence), on-device photo (lightweight classification), triggered interaction (call/message), partner QR/NFC.
  • Anti-abuse: minimum delays, place/time coherence, reasonable manual validation.
  • Privacy: local processing by default, granular opt-in, encryption.

10) Jarvis — Secure Autonomous Agent

  • Plans (calendar), prepares (to-do/brief), summarizes (emails/docs), proposes (events).
  • Connectors: Calendar, Health (HealthKit/Google Fit), Finance (aggregators), Mail, Tasks.
  • Plan → Act → Reflect loop with guardrails (explicit permissions, audit log).

11) Architecture & MVP (Running Code)

  • Frontend: React Native (Expo) mobile + minimal Next.js web.
  • Backend: Supabase/Firebase (auth, Postgres, storage, realtime).
  • AI: LLM (chat), STT/TTS (voice), agent orchestrator.
  • Data: Users, Goals, Quests, Proofs, Scores, Events.
  • Quality: “happy path” < 5 clicks, offline cache (manual validation), network retries, metrics (LLM latency, POA success rate).
  • Hackathon deliverables: runnable README, seed scripts, Postman, 60–90s video, QA checklist.

12) Privacy, GDPR & Safe-Meet

  • Separate consents (GPS, mic/camera, health/finance, matchmaking) — opt-in/opt-out anytime.
  • GDPR: minimization, portability, right to erasure; encryption in transit/at rest.
  • Safe-meet: share live location with a trusted contact, SOS button, safety guidance.
  • Anti-addiction: notification limits; “get off the screen” focus.

13) Market & Targets

  • Core: 18–40 urban self-improvers, casual athletes, young pros, singles.
  • Affinity segments: fitness & wellness, productivity, serious dating, career coaching.
  • Massive TAM (cross-vertical); wedge = POA + Jarvis.

14) Business Model

  • Freemium: basic quests & scores.
  • Premium (€9–15/mo): multi-tool Jarvis, advanced POA, themed programs, evolving avatar, analytics.
  • B2B partnerships: gyms, events, coworking, training (CPA/affiliation).
  • Marketplace: micro-programs (coaches & creators) — rev-share.

15) Go-To-Market

  • PLG + content: 7/21-day challenges, “real > virtual” shorts.
  • Ambassadors: fitness/wellness coaches, therapists, career advisors.
  • Local partnerships: gyms, yoga studios, afterworks, universities.
  • Buddy system: city/campus groups, team challenges.

16) Early Traction (Hackathon → Alpha)

  • Live prototype: chat + voice, quests + 1 POA, simple dashboard.
  • Private alpha: waitlist, pilots with local partners.
  • T1 targets: Day-1 activation > 40%, 3 POA/week/user, D7 > 25%, premium conversion 3–5%.

17) Core KPIs (Steering)

  • Weekly POA/user, avg streak, real/virtual ratio.
  • Day-1 activation, D7/D30 retention, NPS.
  • Premium conversion, ARPU, CAC:LTV.
  • Participation in IRL events.

18) 12-Month Roadmap

  • Q1: FR/EN alpha, robust GPS/photo POA, calendar/health connectors.
  • Q2: Multi-tool Jarvis, premium programs, animated avatar, buddy system.
  • Q3: Public launch FR/EN, national partnerships (fitness/events), meetup safety pack.
  • Q4: Internationalization (EN/ES/DE), program marketplace, partner API.

19) Competitive Landscape & Moat

  • Wellness/productivity: measure but don’t drive action.
  • Dating: optimizes matching, not IRL action.
  • Generic AI coaches: advice without proof of action.
  • AURA moat: proprietary POA dataset, long-term personalization, secure Jarvis, partner network.

20) Impact & ESG

  • Reduces isolation; boosts physical/social activity.
  • Encourages positive mental health via action and connection.
  • Accessibility: voice guidance, high contrast, simple routines.

AURA COMPANION = the AI that makes you live, not scroll.
A life cockpit, an agent that acts, and an ecosystem that brings you back to the real world.
Join us to build the global Proof-of-Action standard.

From scratch