
ObsiFlow is the first 'Truth-to-Transaction' protocol that cryptographically verifies macro-economic news on-chain to trigger automated DeFi execution. The backend provides: Automatic wallet tracking - Track user wallets and discover holdings automatically Sub-wallet discovery - Automatically discover related wallets via Nansen API AI-powered strategies - AI automatically generates trading strategies based on aggregated data Two-phase signals - Immediate signals (no FDC) → Verified signals (with FDC attestation) Multi-source data aggregation - News, market data, whale movements, social sentiment Real-time notifications - WebSocket and Redis Pub/Sub for live updates Knowledge graphs - Visual proof of AI reasoning with interactive graphs Trading bot integration - Subscribe to signals and recommendations via WebSocket/Redis User preferences - Customizable token and network preferences that filter all data, strategies, and AI recommendations FTSO price feeds - Real-time price feeds from Flare Time Series Oracle integrated into knowledge graphs and AI analysis Master knowledge graph - Automated generation linking all whale wallet transactions with news and price data TEE Integration - Deploy AI agent to GCP Confidential Space for verifiable execution with cryptographic attestations

I’ve spent the last 2 months ideating and building the "anti-failure" engine for founders. 🛠️ 90% of startups die. Not because they can't code. But because they build things nobody wants. So I built something to do the heavy lifting. Introducing ProblemCapital. https://problemcapital.vercel.app We don't find customers for your product. We find the product your customers are already screaming for.

ZKLinic is a ZK-native clinical trial recruitment protocol built on Midnight Network. It solves the single most expensive problem in medical research - 85% of clinical trials fail to recruit enough patients on time, not because eligible patients don't exist, but because finding them requires those patients to hand their entire medical record to a pharma company they've never met. Here is exactly where Midnight sits in the architecture. A doctor signs a structured health credential for their patient - diagnosis codes, lab values, medication history - using their on-chain registered wallet. The credential lives on the patient's device. It never touches any server. When the patient applies for a trial, ZKLinic generates a ZK proof on their device that checks their credential against the trial's eligibility criteria. That proof - containing only boolean results per criterion, no clinical data - is posted to Midnight's public ledger. The trial recruiter queries the proof ID and sees exactly one thing: this person is eligible, criterion by criterion, verified by a licensed physician, proven by cryptography. They see nothing else. What is verified on-chain: credential issuer validity, eligibility booleans, proof binding to a specific trial. What stays private: the patient's identity, their exact diagnosis, their lab values, their age, their medication history, every clinical detail that produced those booleans. The verifier is the clinical trial recruiter. What they need to see is a yes or no per criterion, backed by a doctor's signature and a ZK proof they can verify independently on the Midnight explorer. That is all they get. That is all they need. ZKLinic is not a diagnostic app. It is not a medical records platform. It is the trust infrastructure that lets a patient prove they qualify for a trial without becoming a data source for the institution running it.