The chain, as a database you can actually query.

Raw chain data is unusable and your RPC is slow. We build the indexer that follows every block, decodes the events you care about, and keeps a fast database in sync, reorgs and all.

Start a build

Postgres + GraphQLreorg-safe0 behind head

Indexer · Ethereumin sync · 21,450,112

21,450,112TransferUSDC · 1,200→ indexed

21,450,111SwapETH → USDC · 0.05% pool→ indexed

21,450,110MintAcme #4471→ indexed

ingest pipeline0 blocks behind head

Block

#21,450,112

Decode

against the ABI

Postgres

upsert by log id

throughput

1,240 /s

events indexed

4,912,004

reorg-safe

yes

That's the head of the chain, decoded and written as it arrives. Every row is a typed record you could query a second later.

A reorg doesn't corrupt the data.

Naive indexers write the head as if it's final and break the first time the chain reorganizes. Trip one and watch the orphaned rows roll back in a transaction, then re-index on the canonical fork.

events · decoded rowsin sync · head confirmed

blockeventmetastatus

21,450,112SwapETH → USDCindexed

21,450,111TransferUSDC · 1,200indexed

21,450,110Depositvault · 3.2 ETHindexed

21,450,109MintAcme #4471indexed

21,450,108Approvalrouter · maxindexed

Decoded, not raw.

The chain speaks in hashed topics and packed bytes. We decode each log against the ABI into a typed row, so your app reads a transfer, not a wall of hex.

raw log

1topics[0]  xddf252ad1be2c…d5b3ef   // keccak("Transfer(address,address,uint256)")
2topics[1]  x0000…fe41b            // from
3topics[2]  x0000…ac7d2              // to
4data       x000000000000000000000000000000000000000000000000000000000000 b0

decoded row

1Transfer(
2  from:  x9f…e41b,
3  to:    x3a…c7d2,
4  value: 1200000000        // 1,200 USDC (6 decimals)
5)

Then query it like any other database.

Once it's indexed, the chain is just tables. Pick a question, read the SQL, and get an answer in milliseconds, or point your app at the GraphQL layer.

query.sql

1select pool, count(*) as swaps, sum(amount_usd) as volume
2from swaps
3where block_time > now() - interval '24 hours'
4group by pool order by volume desc limit 3;

poolswapsvolume

ETH / USDC5,104$5.1M

WBTC / ETH2,733$2.7M

DAI / USDC1,618$1.6M

3 rows · 12 ms

The chain, as a database you can actually query.

Every block followed, the events you care about decoded to typed rows, reorg-safe, backfilled and live. A Postgres schema and a GraphQL layer, whichever your app speaks.

Start a build The rest of Crypto