Work / Evidence repository
Systems that exist. Numbers that check out.
Every entry answers four questions: what was broken, what we built, where a person stayed in charge, and what changed. Today these are Blocpod’s own systems and public products, labeled as such. Client stories join this page only when a client says we can share them.
Evidence over theater applies to us first.
- BlocpodNFT platform: creator commerce with on-chain provenance and governed paid actionsA creator-commerce and marketplace platform where every mint, listing, sale and licence is a bounded, verifiable action. Released to a funded testnet with a machine-readable release-truth record instead of a launch announcement.343 / 343Backend checks141 / 141Browser checks0Dependency vulnerabilitiesFunded testnet verified 2026-08-14. Release status: not ready. Two gates open and published.
- IssuerOS: governed investor communications where 17 agents recommend and humans decideA multi-tenant operating system for creating, verifying, approving and launching investor-facing communications. Source to claim to content to approval to release, with a hash-chained audit log per organization.17Governed agentsHash-chainedAudit log11Lifecycle steps under governancev0.9.0. Multi-service platform with live-capable SEC/EDGAR and GitHub adapters; other integrations explicitly labeled mock.
- FounderOS on Meridian: a governed executive office running on verified contextMissions, specialist agents, compiled context packs and approval boundaries for a founder's real workload. External actions are draft-first, memory is gated, and every fact carries its source.3Authority classes~1 to 3k tokensContext pack budget0Auto-send actionsPilot-stage internal system. Research prototype and local runtime; production auth, live source integrations and the blind evaluation harness are open work.
- Blocpod Workforce: from a job description to a governed AI worker with budgets, approvals and auditDescribe the work in plain language. The platform specifies the job, designs candidate agents, auditions them on the same scenarios, and deploys the hire under least-privilege tools, hard spending limits and approval gates.16Lifecycle stagesPer callCost attributionSellable demo. Full lifecycle implemented and resumable; every model and tool call metered to the millicent.
- Bitcoin AI Provenance: signed receipts for what an agent did, anchored where nobody can edit themAn open-source reference implementation for provenance receipts and bounded autonomous-agent actions. Ed25519 identities, cycle-checked lineage, signed agent policies with spend and action limits, Merkle batching, Bitcoin anchoring.0Runtime dependenciesApache 2.0LicenceUnit, integration, adversarial, determinism, releaseTest gauntletsPublic, Apache 2.0, CI green. Deliberately not a rollup, decentralized storage network, ZKML prover or BitVM bridge; those are documented as research tracks.
- Velvet: a consumer preference platform with a governed AI recommender and a partner APIA taste-and-preference platform built for a live consumer event: anonymous sessions, server-owned consent, an inventory-aware recommender with an AI rerank behind cost ceilings and timeouts, Brand Studio analytics, and a versioned partner API.46 unit + 3 end-to-endAutomated tests43 recipesCataloguev1Partner APIBuilt and tested for a live consumer event (July 2026). Partner API v1 shipped; second product (TequilaTown) integrates through it.
Your workflow could be the next entry
Bring us the bottleneck.
Client work appears here only with permission and only with verified numbers. Bring us the bottleneck and we will tell you what the evidence would look like before we build anything.