Megalo

1. Inspiration
2. Legacy creates data silos → poor visibility, higher breach risks, and compliance headaches (harder to apply consistent privacy controls or respond to data subject rights).
3. Maintenance costs are underestimated by 70-80%; total ownership is often 3–4x higher.

4. Talent gap: Retiring mainframe experts make support unsustainable 2026 Banking and Capital Markets Outlook (Oct 2025): AI ambitions are "throttled by brittle and fragmented data infrastructure" and "outdated legacy systems." Siloed data and legacy prevent meaningful outcomes in compliance, cyber defense, and innovation. ( Report By Deloitte )

5. What it does A borrower connects their financial documents through regulated open banking APIs or direct upload. The data never leaves their device or secure environment. An AI reads the documents, extracts only the signals a credit decision needs, and feeds them into a verification process. Lenders receive a single output: a credit grade and a loan offer — nothing else. No raw income figures. No account numbers. No document copies stored anywhere.

6. How we built it LayerTechnologyPrivacy proofMidnight blockchain (Compact ZK circuits)AI — document parsingClaude API (Anthropic)AI — rate pricingClaude API (Anthropic)AI — score explanationClaude API (Anthropic)AI — cohort analysisClaude API (Anthropic)Data ingestionPlaid (US) · Open Banking API (UK/EU PSD2) · direct uploadIdentity attestationExisting bank KYC via open banking OAuthFrontendNext.js + TailwindOn-chain stateMidnight shielded contracts

7. Challenges we ran into Deployment on pre prod not doing well cuz network been down and bad UX for the deployement in the local

8. Accomplishments that we're proud of This is the our first agentic privacy product

9. What we learned We learned more about the banking and privacy usecases and been engaged with connection for the real problem validation

Canva link: https://canva.link/cwjr77o92s3o3jz

Built With

• claude
• midnight
• nextjs