Futurizer

Futurizer is a decentralised prediction market platform that makes collective intelligence accessible to anyone, not just crypto-native users. Think of it as a way to turn any question about the future into a tradeable market — resolved automatically by AI, settled trustlessly on-chain.
We built three distinct modes:
Public markets — Anyone can create a market via Telegram in plain English. No wallet setup, no gas fees. The platform handles everything: market creation goes through a Chainlink CRE workflow, and when it's time to resolve, three AI investigators (Gemini and DeepSeek) independently research the outcome and vote. Majority wins. The result is signed by Chainlink's DON and written on-chain — no human oracle, no single point of trust.
Private markets — Companies can run internal prediction markets on sensitive questions like "Will we ship this feature by Q3?" or "Will we hit our sales target?" without exposing any data publicly. The question hash goes on-chain, but the question text and resolution source stay off-chain. Chainlink's ConfidentialHTTP fetches the company's private resolution API inside the DON's trusted enclave — API keys are injected from a vault and never visible to anyone. This gives organisations a powerful forecasting tool with full confidentiality.
LMSR AMM — New prediction markets suffer from thin liquidity: if only a few people participate, there's no one to trade against. We solve this with a Logarithmic Market Scoring Rule AMM. The market creator seeds liquidity upfront, and the AMM acts as the always-available counterparty. Prices adjust automatically with demand. Even a market with three participants is fully tradeable from day one.
The result is a prediction market platform that works for everyday Telegram users, enterprise teams, and niche communities alike — all backed by Chainlink for verifiable, trustless settlement.

We had four people, each owning a clear slice of the stack:
Akari — CEO and project lead, keeping the team focused and aligned, redirecting priorities when needed, and driving the marketing and presentation side of the project.  
Owen — AI resolution agent (multi-model parallel investigators with majority-vote judge), private market Telegram commands, CRE private create/settle workflows, and the full shielded prediction user flow including the ACE Vault integration.
Silverlynx — Smart contracts (Futurizer, FuturizerLMSR, Treasury, FUSD), LMSR AMM logic, and the public create/settle CRE workflows.
Richard — Integration, end-to-end testing, and making sure all three layers — contracts, API, and bot — ran together reliably.
Clear ownership let us move fast without stepping on each other. The API server (Express + Telegraf) acts as the glue layer — it receives Telegram commands, triggers the right CRE workflow via cre workflow simulate --broadcast, and routes results back to users.