Blockchain data infrastructure and AI agent payments
Edge & Node
Edge & Node builds blockchain data infrastructure and payment controls for AI agents, alongside advisory services for institutions adopting digital assets. Its flagship Amp turns activity across multiple blockchains into structured datasets accessible through SQL. Banks, stablecoin issuers, and analytics teams can use it to analyze treasury flows and trace results back to blockchain records through cryptographic provenance. Amp supports cloud, on-premises, and hybrid deployment. ampersend gives agent developers wallets, spending limits, approved-counterparty lists, and transaction monitoring for paid services.
The company grew out of Graph Protocol, Inc., whose transition into Edge & Node was announced in 2020. Its team created The Graph, an indexing protocol that makes blockchain data accessible through GraphQL, and continues contributing to its open indexing infrastructure. Edge & Node is led by CEO Rodrigo Coelho, an early team member who previously led community efforts and managed operations. Amp extends this engineering background into institutional data workflows, preserving a verifiable trail from source records to queries.
ampersend separates an agent’s ability to request payments from authority to spend. Its smart-account architecture requires agent signatures to be co-signed by a server that enforces user-defined policies. Agents can buy services and accept payments through x402, which embeds stablecoin payments in HTTP requests. A routing integration with Amazon Bedrock AgentCore Payments also lets agents access multiple model providers and pay per request through one integration, with session budgets enforced by AgentCore.
Explore the recordings
The supplied Edge & Node archive contains one recording, Agent Spending Without Controls, presented by Rodrigo Coelho and Pranav Maheshwari. The paths below guide readers through its arguments and demonstrations. Product capabilities, usage figures, and historical accounts are claims made in the recording; this guide does not verify their status today.
Start with the case for controlling agent spending
Watch Agent Spending Without Controls for the presenters’ argument that agents making continuous payments require different controls from human purchasers. Coelho discusses identity, sanctions screening, spending limits, and policy enforcement as enterprise deployment requirements. This path is useful for understanding the proposed control model and its rationale.
Follow the payment infrastructure argument into the tool demos
In the same recording, Coelho connects Ampersand to his account of The Graph’s query scale, its 2021 micropayment system, and later work around Coinbase’s x402 specification. Maheshwari then demonstrates a skill file providing paid MCP-tool access and payments, followed by a Shopify UCP gift-ordering example. Use this path to connect the infrastructure discussion to the demonstrated agent workflow. The gift demo’s narrated budget and charge amounts conflict, so it cannot establish that the purchase stayed within the stated limit.
Inspect the screening simulation
The final simulation in Agent Spending Without Controls compares scraping micropayments from a normal wallet and a wallet represented as sanctioned. Both authorize with screening disabled; enabling the demonstrated TRM-backed screening rejects transactions from the blocklisted wallet. This illustrates the presenters’ proposed enforcement mechanism within the simulation, without establishing comprehensive compliance coverage or current production behavior.
1 talk
Newest first
2 speakers at AIE
Affiliations reflect their AIE appearances, not necessarily current employment.