Labeling

Inflecta processes on-chain data using a multi-layer labeling and aggregation framework. Each layer captures a different dimension of a wallet’s economic role and enables analysis at multiple levels of abstraction.

This framework forms the foundation for distinguishing economically meaningful flows from mechanical or operational noise.

1. Wallet business logic

Wallets are first classified according to their observable on-chain economic behavior.

This classification is derived from repeated behavioral patterns such as interaction counterparties, transaction structure, turnover, balance management, and historical activity across chains and protocols.

Examples include:

• Hot/cold wallets

• Liquid funds

• Asset Managers

• OTC desks

• Market makers, dealers

• Arb/MEV bots

• Incentives distribution, vesting contracts

• Lending pools, DEX pools, and other DeFi protocol contracts

• Smart Money (in Inflecta this label is not behavioral and is assigned only to known industry participants such as founders, fund managers, and other recognized operators).

• Individual, Angel, and Farmer are behavioral labels describing the primary role of the wallet.

• Unknown Buy Side (directional exposure is observed but insufficient information exists to classify the wallet further. This label is often assigned to unknown CEX or DEX accumulators with limited transaction history).

2. Relationship to the project

Separately from global behavior, wallets and entities may have a project-specific role, including (but not limited to):

• Treasury wallet

• Team wallet

• Contributor distribution receiver

• Investor distribution receiver

• Incentives

Insiders and project-related distribution wallets are tracked across multiple hops, provided the ultimate beneficiaries are assessed as remaining economically unchanged through behavioral and entity-level analysis. This allows Inflecta to reconstruct the true economic ownership path of token allocations beyond the original distribution.

3. Entity Attribution

Where sufficient evidence exists, wallets are clustered into entities representing known organizations or individuals.

Examples include:

• Pantera Capital

• Binance

• Wintermute

Entity attribution relies on a combination of behavioral similarity & entity business logic applied through heuristics and pattern analysis, as well as known infrastructure.

If behavioral clustering and business logic are clear but attribution is uncertain, entities are represented using stable placeholders (for example, Liquid Fund 16).

Entity attribution is probabilistic. Where applicable, Inflecta provides a confidence score reflecting the strength of the attribution.