Bloodline v0.9 · pre-seed
mybloodline.in ricky@mybloodline.in
Technical documentation

Infrastructure for what happens to your assets when you can't manage them yourself.

Bloodline is an AI-native platform for discovering, verifying, protecting, and recovering financial and digital assets on behalf of families. It combines agentic AI, blockchain attestation, secure storage, and identity verification into infrastructure that heirs can actually use — not just a legal document describing what should happen.

Status Pre-seed · building Phase 1
Model AI-agnostic orchestration
Chain EVM-compatible, selective anchoring
Contact ricky@mybloodline.in

The problem

Traditional financial systems assume the account holder stays available to manage their own assets. That assumption fails the moment someone is incapacitated or dies.

Traditional assets

Forgotten bank accounts, insurance claims never filed, dormant investments, and pension benefits that go unclaimed because no one knew where to look.

Web3 assets

Self-custodied assets can't be recovered without private keys or seed phrases — and no institution can reset or recover them on a beneficiary's behalf.

Bloodline builds the infrastructure layer that minimizes this risk while preserving user ownership and privacy — for both worlds, on one platform.

Design principles

User ownership of all discovered data
Privacy by default, not by configuration
AI-first automation across recovery workflows
Institution interoperability over walled gardens
Blockchain used only where trust benefits
Human oversight required for high-risk actions
System design

Architecture

A single orchestration layer routes every request through specialized AI agents, which in turn reach into institutions, chains, and storage — never the other way around.

USER WEB / MOBILE CLIENT API GATEWAY — REST + GRAPHQL AI ORCHESTRATOR Agent workflow engine · memory · tool use DISCOVERY AGENTS RECOVERY AGENTS VERIFICATION AGENTS EXTERNAL SURFACE Institution APIs Blockchain Document Storage Identity Providers Notifications Banks · brokerages · exchanges · wallets · eKYC · email · SMS · push
4.1 — Asset discovery engine

Continuously maps a user's financial footprint across banks, brokerages, insurance, pensions, exchanges, wallets, NFTs, and DeFi protocols.

  • API integrations
  • Uploaded documents & statements
  • Wallet connections
  • Email parsing (planned)
4.2 — AI recovery copilot

Guides beneficiaries through institution-specific recovery workflows end to end.

  • Document understanding & eligibility checks
  • Checklist generation & form completion
  • Follow-up reminders & progress tracking
4.3 — Proof-of-life engine

Periodically verifies user activity without exposing unnecessary personal information, then decides whether continuity mode should trigger.

  • App activity & wallet signatures
  • Biometric / government ID checks
  • Trusted contacts & institutional confirmation
4.4 — Digital vault

Encrypted storage for the documents a recovery actually depends on.

  • Wills & beneficiary instructions
  • Insurance & identity documents
  • Wallet metadata & recovery notes

Lifecycle

01User creates an account and verifies identity
02Assets are discovered and documents uploaded
03AI organizes the inventory and beneficiaries are configured
04Consent is recorded and Proof-of-Life monitoring begins
05A trigger event is detected and recovery workflow initiates
06AI assists beneficiaries through to recovery completion
Agentic layer

AI Agents

Four specialized agents collaborate under the orchestrator to complete recovery workflows that would otherwise require a human caseworker at every institution.

Discovery agent
  • Identify relevant institutions
  • Classify uploaded documents
  • Extract structured information
  • Detect missing or likely assets
Verification agent
  • Identity verification
  • Beneficiary validation
  • Document consistency checks
  • Fraud detection
Recovery agent
  • Generate institution-specific workflows
  • Prepare documentation
  • Answer beneficiary questions
  • Track recovery progress
Planning agent
  • Prioritize recovery tasks
  • Estimate complexity
  • Recommend next actions
  • Generate personalized continuity plans

AI stack

The current architecture is model-agnostic by design — no single LLM vendor is load-bearing.

Gemini GPT Claude

Supporting infrastructure:

RAG OCR Document embeddings Vector databases Workflow orchestration Function calling Tool use

Agent memory persists context across long-running recovery processes, which can span weeks across multiple institutions.

Trust layer

Security

Security follows a zero-trust approach end to end. No plaintext recovery secrets are ever exposed to Bloodline operators.

AES-256 encryption at rest
TLS encryption in transit
Role-based access control
Multi-factor authentication
Hardware-backed key management
Encrypted backups
Consent-based access at every step
No operator access to plaintext secrets

Privacy model

Users retain ownership of their data. Bloodline operates strictly under explicit, revocable consent — access permissions are granular by default, never blanket.

GDPR ADGM Data Protection Regulations DIFC Data Protection SOC 2 ISO 27001

Compliance targets — future state, not current certification.

Trust layer

Blockchain

Blockchain is used selectively — for immutability and verification, never as a place to store sensitive personal data.

Anchored on-chain
  • Document hashes
  • Beneficiary approvals
  • Consent records
  • Workflow attestations
  • Audit trails
Kept off-chain
  • Identity documents
  • Wallet metadata & seed material
  • Wills & beneficiary instructions
  • Any personally identifiable information

Candidate infrastructure

Polygon Base Ethereum L2 Consortium chain IPFS
Build with Bloodline

APIs — planned

REST and GraphQL surfaces are planned for Phase 2. The groupings below reflect the integration surface, not a stable contract yet.

Financial institutions
  • Banks
  • Brokerages
  • Insurance companies
Digital assets
  • Wallet providers
  • Exchanges
  • Custodians
Identity
  • Government eKYC
  • Identity providers
Notifications
  • Email
  • SMS
  • Push

Technology stack

Frontend
ReactNext.js
Backend
PythonFastAPINode.js
AI
GeminiOpenAILangGraphLangChainMCP (planned)
Storage
PostgreSQLRedisVector DBObject storage
Blockchain
EVM-compatibleIPFS
Infrastructure
DockerKubernetesCloud-native
Where this is going

Roadmap

Four phases, moving from financial-asset recovery today toward an autonomous, cross-border AI family office.

Phase 1 · current

Foundation

  • Financial asset discovery
  • Recovery Copilot
  • Secure vault
Phase 2

Digital assets

  • Digital asset discovery
  • Wallet integrations
  • Institutional APIs
  • AI recovery automation
Phase 3

Autonomy

  • Family wealth graph
  • Cross-border recovery
  • Autonomous AI agents
Phase 4

Wealth continuity platform

  • Digital asset-backed lending
  • Estate financing
  • AI family office
Reference

FAQ

What is Bloodline, in one sentence?

Infrastructure that lets heirs actually discover and recover a family's financial and digital assets — not just legal paperwork describing what should happen to them.

How is this different from estate planning software?

Estate planning software focuses on documenting intent — wills, trusts, instructions. Bloodline builds the operational layer underneath that: continuous asset discovery, Proof-of-Life monitoring, and AI-assisted recovery workflows that actually execute across institutions and wallets.

Does Bloodline ever hold my private keys or seed phrases?

No. Wallet metadata and recovery instructions can be stored in the encrypted vault, but private keys are never transmitted without explicit user consent, and no plaintext recovery secret is exposed to Bloodline operators.

What triggers a recovery workflow?

The Proof-of-Life engine periodically checks activity signals — app usage, wallet signatures, biometric or ID confirmation, trusted contacts — and determines whether the account should stay in normal monitoring or transition into continuity mode.

Why use blockchain at all if sensitive data stays off-chain?

Chains are used for what they're actually good at: making consent records, approvals, and audit trails tamper-evident. Anything personally identifying stays off-chain by design.

Which LLMs power the agents?

The orchestration layer is model-agnostic. Gemini, GPT, and Claude are all candidate models, selected per task rather than locked to a single vendor.

What's actually built today versus planned?

Current work is focused on Phase 1: asset discovery, AI-assisted recovery, secure beneficiary management, document intelligence, and institutional workflow orchestration. APIs, deeper wallet integrations, and autonomous agents are Phase 2 and beyond.