700M+ labeled addresses. 27 chains. One query.
Classify wallets, track whale movements, segment users by behavior, and score airdrop eligibility. All in SQL, all crosschain, all with temporal context baked in.
01 / What you can do
Seven use cases you can start running today. Each one maps to real Flipside tables you can query in minutes.
Key question:
"Our protocol shows 50K MAUs — how many are actual humans?"
Key question:
"Are the top 100 holders accumulating or distributing?"
Key question:
"What percentage of our users are DeFi power users vs. casual traders?"
Key question:
"How do we reward real contributors and filter out sybil farmers?"
Key question:
"Where are our new users bridging from, and do they stick around?"
Key question:
"What does this wallet hold across chains, and how is it changing?"
02 / The data layer
Identity labels for who. Balance snapshots for what. Behavioral profiles for how. Same SQL, every chain.
Explore label coverage700M+ addresses classified across 27 chains
Join any transaction or balance query with dim_labels to instantly know if the counterparty is an exchange, protocol, whale, bot, or fund. dim_tags adds temporal context — when a classification was assigned and when it expired.
-- Classify 30-day active addresses
SELECT r.from_address, r.blockchain,
r.tx_count, l.label_type,
CASE
WHEN l.label_type IS NOT NULL THEN 'KNOWN_ENTITY'
WHEN r.tx_count > 5000 THEN 'LIKELY_BOT'
WHEN r.total_fees > 0 THEN 'ORGANIC_USER'
END AS user_class
FROM crosschain.evm.fact_transactions r
LEFT JOIN crosschain.core.dim_labels l
ON r.from_address = l.addressDaily token balances with USD values
Track what every address holds, every day. Compare balances over 30-day windows to spot accumulation and distribution patterns. Available for ERC-20 tokens across EVM chains, Solana SPL tokens, Aptos, and NEAR.
-- Top holders with 30-day change signal
SELECT address, symbol, balance_usd,
balance_usd - COALESCE(prior.balance_usd, 0)
AS usd_change_30d,
CASE
WHEN balance_usd > prior.balance_usd * 1.25
THEN 'ACCUMULATING'
WHEN balance_usd < prior.balance_usd * 0.75
THEN 'DISTRIBUTING'
ELSE 'STABLE'
END AS whale_signal
FROM crosschain.balances.ez_balances_erc20_daily
WHERE balance_usd >= 100000One row per address, full behavioral summary
Skip the multi-join queries. ez_{chain}_address_metrics aggregates every address's activity across DeFi swaps, NFT purchases, governance votes, staking actions, and total transactions into a single row. Available across 11 chains.
-- Segment users by behavior
SELECT address,
CASE
WHEN total_dex_swaps > 100
THEN 'DEFI_POWER_USER'
WHEN total_nft_purchases > 20
THEN 'NFT_COLLECTOR'
WHEN total_governance_votes > 5
THEN 'GOVERNANCE_PARTICIPANT'
ELSE 'GENERAL_USER'
END AS user_segment
FROM crosschain.chain_stats.ez_ethereum_address_metrics
WHERE total_txn_count > 003 / Agent intelligence
AI agents run wallet queries on a schedule, detect behavior changes, and deliver findings to Slack or email. Set it once, get alerts when something moves.
Get pinged when a top-100 holder starts distributing, when a known fund takes a new position, or when large amounts move to exchange deposit addresses.
Track the ratio of organic vs. bot activity over time. Know immediately when sybil farming spikes or when your quality user percentage drops below a threshold you set.
Weekly reports showing how many users returned, how many churned, and which segments are growing. Broken down by user type so you know if you're losing power users or casual ones.
Monitor bridge inflows and outflows for your chain. Agents flag when a competitor chain starts pulling your active users or when a new chain sends a wave of bridged capital.
04 / Who uses this
Measure real users, design fair airdrops, and track post-distribution retention.
Track whale accumulation, holder concentration, and crosschain capital flows.
Understand where users come from, whether they stay, and what brings them back.
Segment users by behavior, personalize experiences, and build wallet analytics into your product.
700M+
Labeled addresses
27
Chains with label coverage
11
Chains with address profiles
Daily
Balance snapshot frequency
Flipside combines algorithmic detection with human curation across 27 chains. The dim_labels table maps addresses to known entities (exchanges, protocols, whales, bots) while dim_tags tracks temporal classifications — so you can see that an address was a whale from March to June, not just that it's a whale now.
Yes. Join transaction data with dim_labels to filter out known entities, then use quality scoring (Flipside score of 4 or higher) from ez_{chain}_address_metrics to isolate organic activity. You can also analyze transaction frequency patterns to flag addresses with bot-like consistency — high tx count plus 25+ active days per month is a strong signal.
Daily balance snapshots via ez_balances_erc20_daily across EVM and Solana chains, CEX flow tables for large exchange deposits and withdrawals, dim_labels for known fund and institution addresses, and bridge activity tables for crosschain capital migration. You can track the top 500 holders of any token with 30-day accumulation and distribution signals.
ez_{chain}_address_metrics provides a single-row behavioral summary per address covering DeFi swaps, NFT purchases, governance votes, staking actions, and total transaction counts. Available across 11 chains. No multi-join queries needed — just filter and segment directly.
Yes. Build multi-dimensional engagement scores from ez_dex_swaps, ez_lending_deposits, ez_bridge_activity, and fact_transactions. Score addresses on consistency, DeFi usage, lending, crosschain activity, and tenure. Use dim_labels to automatically exclude known exchanges, protocols, and bots from eligibility.
Flipside's crosschain schema normalizes data across 20+ chains with standardized column names. Bridge activity tables track source and destination chains, so you can see where users come from, how much they bridge, and whether they stay active after arriving. dim_labels unifies address identity across all chains.
Get Started
700M+ labeled addresses, 27 chains, pre-computed behavioral profiles. Tell us who you're trying to understand and we'll show you the data.