Platform Overview

One Platform. Three Phases.
Compounding Outcomes.

The growth marketing system that compounds. Built on your first-party audience and interest graph. Composable across every channel you spend on.

Try FreeBook a Demo
Phase 1 · Foundation

Your audience and interest graph.

Read your warehouse, Snowflake, Databricks, BigQuery, without moving data. Build immutable identities anchored in your data. Layer in real-time signals across web, app, transactions, partners, and intent. Score every customer continuously on FIRE: Fit, Intent, Recency, Engagement.

Audience

Immutable IDs anchored in your warehouse.

  • Immutable IDs (HMAC-SHA-256), never expires, never rented
  • Identity resolution across people, accounts, devices, emails
  • Activity schema, canonical, time-aware event model
  • AI lookalikes generated from your best buyers
Unified identity graph across people, accounts, devices, anchored in the warehouse

Interest Graph

1P plus 2P plus 3P signals unified, time-aware.

  • Real-time ingestion from web, product, ads, cart, intent tools
  • FIRE scoring updated continuously: Fit, Intent, Recency, Engagement
  • Consent-based 2P/3P enrichment
  • Time-aware: interest decays, recency matters, freshness compounds
Time-aware interest graph compounding signals from owned and partner sources
Phase 2 · Activation

Every channel. Multi-channel ads. Match-rate boosting built in.

Push prioritized audiences and next-best-actions to whatever marketing or ad tools you already use. 200+ integrations. Bidirectional sync. Match rates from 20 to 40% to 70 to 90% on Meta, Google, LinkedIn, Reddit. Channel-agnostic, your channels stay your channels.

Activations

Every channel, matched plus synced.

  • 200+ integrations: paid, owned, retail media, agent surfaces
  • Match-rate optimization plus suppression
  • Bidirectional sync, outcomes flow back to Optimization
  • Channel-agnostic, no vendor lock-in
Decision layer routing audiences to paid, owned, retail media, and agent surfaces
Phase 3 · Optimization

Every cycle smarter. This is the moat.

Causal AI plus the Decisioning Waterfall plus FIRE scoring decide who, when, and what. iWorkers, role-based digital twins, execute. Domain-adapted Small Language Models (SLMs) run inside your warehouse: zero egress, low inference cost. Every decision is logged as a Decision Trace, auditable, explainable, defensible. Outcomes feed back. Segments sharpen. Agents improve. The platform compounds.

Decisions

Causal AI plus role-based agents decide who, when, and what.

  • Decision Engine: Decisioning Waterfall plus Causal AI plus propensity models
  • iWorkers: CMO orchestrator plus 26 specialized agents (15 D2C plus 11 B2B), 100+ tools
  • Intelligence: model orchestration, signal fusion, prediction surfaces
  • Skill Packs: 104 modular skill packs that iWorkers compose
  • Decision Traces: every decision logged, signal to iWorker to FIRE to action to outcome
Decision engine selecting next-best-action and routing to channels

Learning

Every decision written back. Outcome memory refines the loop.

  • Outcome memory, every decision and its result, written back to your warehouse
  • Performance pillar, analytics view of what is compounding
  • Feedback loops, auto-update audience profiles, signal thresholds, agent prompts
  • Compounding memory, the brand-specific learning asset competitors cannot replicate
Outcomes feeding back into the audience and interest graph as the system compounds

Integrates with the data clouds, CRMs, and ad platforms your team already runs on

Snowflake logoDatabricks logoGoogle Cloud logoReltio logoHubSpot logoLinkedIn Ads logoGoogle Ads logoMeta Ads logodbt Labs logoSnowflake logoDatabricks logoGoogle Cloud logoReltio logoHubSpot logoLinkedIn Ads logoGoogle Ads logoMeta Ads logodbt Labs logo

Features FAQs

How the Composable Decision OS works under the hood.

No. iCustomer is not a CDP and does not store or move data. It is not a campaign tool, it does not run campaigns or create creatives. iCustomer is the Composable Decision OS, the decision context layer that reads your data, builds audience and interest context, decides who to target, when, what, and in which channel, and activates that audience in the tools you already use.
FIRE is iCustomer's account and customer scoring framework. F is Fit (how well they match your ideal profile), I is Intent (buying signals right now), R is Recency (how recently they engaged), and E is Engagement (depth and quality of interaction). Every customer and account gets a composite FIRE score, updated continuously from your data signals.
A Decision Trace is iCustomer's audit record for every targeting decision. It logs what signal triggered the decision, which iWorker acted, what the FIRE score was, and what action was recommended. Decision Traces make every decision explainable and defensible, for your team, your CMO, and your board.
iWorkers are role-based digital twins, each assigned a specific job in your marketing organization. They run on an always-on OODA Loop and a proprietary Decisioning Waterfall, with causal AI and compounding memory. Every decision is grounded in your actual data, signals, ICP, and campaign history, making output actionable rather than generic. Today: 1 CMO orchestrator plus 26 specialized agents (15 D2C plus 11 B2B), 100+ tools, 104 Skill Packs.
Same product, two entry gates. Audience Loop is the SMB / mid-market gate, self-serve, credit-based, sheet-native, fast install, embedded versions of all five modules. Composable Decision OS is the enterprise gate, sales-led, AUM-tiered, full Decisioning Waterfall, full Causal AI, full iWorker depth. Same shape underneath. ~30% of Audience Loop customers expand to Composable Decision OS within 12 months.
No. iCustomer runs warehouse-native, on Snowflake, Databricks, BigQuery. Zero data movement. Zero egress. Your warehouse stays the system of record. iCustomer leverages your existing data cloud, warehouse, and tools and helps you optimize them in a headless, composable, AI-native tech stack, no migration, no rip-and-replace.

Turn your data
into outcomes

Start free with Audience Loop. Or talk to us about a Composable Decision OS pilot on your warehouse.

Try FreeBook a Demo