Rodrigo Coelho
1.1K posts
Rodrigo Coelho
@rodventures
CEO @edgeandnode | Founding Team of The Graph @graphprotocol | Creators of Amp and ampersend @ampersend_ai
We are heading to Brazil 🇧🇷 Our CEO @rodventures is speaking at @mmerge_io in Sao Paulo on March 19th at 1:10PM on the panel "From Blockchains to Deployment to Compliance: The Infrastructure Institutions Need", alongside @jeffbeltrao from @Chainlinklabs and @daluca from @OpenZeppelin. If you're building infrastructure for institutional adoption, this is the conversation to catch. Come find the Edge & Node team there.
We are heading to Brazil 🇧🇷 Our CEO @rodventures is speaking at @mmerge_io in Sao Paulo on March 19th at 1:10PM on the panel "From Blockchains to Deployment to Compliance: The Infrastructure Institutions Need", alongside @jeffbeltrao from @Chainlinklabs and @daluca from @OpenZeppelin. If you're building infrastructure for institutional adoption, this is the conversation to catch. Come find the Edge & Node team there.
There are some pretty wild downstream effects in a world with trillions of agents using the internet and software. One very big one is what happens with agents with budgets and wallets. There are lots of business models that never ended up working out for the human-based internet that all of a sudden start to make economic sense in an agent-based internet. Think of all the proprietary data and research that’s sitting out there right now behind a paywall that a human will never run into. Finance data, medical research, and so on. Most people won’t sign up for a $100 or $1000 subscription for information they need infrequently. The cost is too high. Equally, micropayments for this data rarely worked at scale because the volume was too low to matter. However, now an agent can have a budget for a specific set of research it’s doing, and the agent might pay $0.1 or $1 to access it in a workflow. And now that data may be relevant in 1,000X’s more use-cases than it was before. Similarly, there are many APIs and tools out there on the web that don’t make sense to have a subscription for, but now an agent may interact with for a specific exchange, and it could cost $.01 or $0.1 per transaction. All of a sudden new kinds of software can get built and monetized that would have been uneconomical before. Some new form of commercial open source, essentially. Obviously lots of infrastructure and agreement across the industry is needed for this -and getting discovered by the agent is going to be a whole new class of search and discovery problem- but there are so many potentially interesting new scenarios here.
Very soon there are going to be more AI agents than humans making transactions. They can’t open a bank account, but they can own a crypto wallet. Think about it.
Amp's benchmarks are mind-blowing. Here are a few to show the innovation this product brings to blockchain data access. ~100 ms latency at chain head. This is the time between a new block appearing on the network and Amp having that data available for streaming SQL queries. For context, Ethereum's block time is 12 seconds, so Amp serves verified data within the first 1% of the next block's window. 4,000,000 events decoded per second. This is the throughput of Amp's evm_decode UDF, processing raw log data into structured, typed output. You can decode every Uniswap swap, ERC-20 transfer, and governance vote across an entire chain's history at this rate. 5 seconds to query all block hashes since genesis. Query: SELECT hash FROM "edgeandnode/ethereum_mainnet".blocks. Every block hash from block 0 to the current head, returned in 5 seconds—thanks to columnar Parquet storage with block-level partitioning. 35 seconds to aggregate all transactions since genesis. Query: SELECT COUNT(*), value > 0 AS has_value FROM "edgeandnode/ethereum_mainnet".transactions GROUP BY 2. Billions of transactions, aggregated with a GROUP BY, in 35 seconds. These numbers come from Amp running on the Apache DataFusion query engine with Apache Arrow as the in-memory format. DataFusion's vectorized execution across columnar data gives Amp the same query optimizations that power the broader analytics ecosystem, not a bespoke blockchain-specific engine. Amp is set to become the new standard for blockchain data access. Check it out: github.com/edgeandnode/amp
The mental model most people have is "AI helps humans do tasks." That’s no longer the case at @Coinbase. Their model is "AI does tasks, humans set guardrails." And they’re using crypto to unlock full agent autonomy.
a $47k incident: two ai agents stuck in a recursive loop for 11 days. no breach, no exploit. just agents with unbounded api keys and no infrastructure-level enforcement to stop them. we built the demo that makes this failure mode impossible (📺 watch the video) the ampersend x @BlockRunAI agentops demo shows a buyer agent and seller agent communicating over a2a, with every llm inference call routed through blockrun via the x402 payment protocol. the seller pays blockrun per request using usdc on base. ampersend's wallet policy enforces the daily spending limit at the wallet layer, not in application code. when the limit is hit, the treasurer refuses to sign. the agent is economically dead. the loop stops. total spend: exactly the daily limit you configured. this matters because step limits and token caps fail under composition. agent a calls agent b which calls agent c. each one self-reports staying within budget. the bug that causes the loop also breaks the counter. anything that depends on the agent's own logic to limit its spend will fail in exactly the scenarios where limits matter most. the architecture here moves budget enforcement out of the trust domain of the agent entirely. the wallet is the circuit breaker. built with ampersend + blockrun + x402. testnet transactions on base sepolia, mainnet-ready. check out the repo ⬇️
