Audience

@absolutejs/audiencev0.0.4alphaCommerce & Growth

Psychographic inference, affinity profiles, and measured audience-overlap scores on your own AI provider.

In-house audience and affinity intelligence — the ownable parts of what an audience-intelligence SaaS sells, run on your own provider and embedding model. Every primitive is a pure function of typed input plus an injected AI call, and affinity work also takes an injected embed call, so the package never picks a provider or an embedding model. Pairs naturally with @absolutejs/partnership, which follows the same injection pattern.

#Installation

BASH

bun add @absolutejs/audience

#Capabilities

Overview

In-house audience & affinity intelligence for AbsoluteJS apps — the ownable parts of what an audience-intelligence SaaS sells, run on your own provider and embedding model.

inferPsychographics — communication style, values, and motivations of a

person or brand from public signals (the "how to approach them" layer).

Show 6 more

profileAffinity — a structured interest/brand affinity profile, embedded

into a vector.

affinityOverlap — a measured audience-overlap score between two profiles

(cosine of their embedded affinities) plus shared topics/brands. Pure, no model call.

cosineSimilarity — the underlying vector math, exported.

Like @absolutejs/partnership, every primitive is a pure function of typed input plus an injected AI call. Affinity work also takes an injected embed call, so the package never picks a provider or an embedding model — your app supplies both, with its own metering and caching.

Measuring audience overlap

For a cheap inline path (no affinity extraction), embed two short descriptor strings yourself and call cosineSimilarity directly.

Psychographic inference

inferPsychographics derives the communication style, values, and motivations of a person or brand from public signals — the "how to approach them" layer.

Affinity profiles

profileAffinity extracts a structured interest/brand affinity profile and embeds it into a vector via your injected embed call.

Measured audience overlap

affinityOverlap returns a measured audience-overlap score (cosine of two embedded affinity profiles) plus shared topics and brands — pure, no model call.

Provider and embedding injected

Your app supplies generateObject and embed through AudienceContext, keeping provider choice, metering, and caching in your own code.

Vector math exported

cosineSimilarity is exported directly for a cheap inline path: embed two short descriptor strings yourself and compare.

Outcomes

What you can build

Overview

In-house audience & affinity intelligence for AbsoluteJS apps — the ownable parts of what an audience-intelligence SaaS sells, run on your own provider and embedding model.

Measuring audience overlap

For a cheap inline path (no affinity extraction), embed two short descriptor strings yourself and call cosineSimilarity directly.

Hardening checklist

Production guidance

Make every external boundary explicitPin the deployed @absolutejs/audience version, replace example or memory-backed dependencies with durable implementations, bound external calls, protect credentials, and emit enough evidence to retry or recover safely.

Follow in order

Troubleshooting path

Trace from the first failed boundary

Reproduce the smallest canonical @absolutejs/audience example, confirm the supported entry point and version in the API explorer, then inspect the first boundary that did not produce its documented result.

#Wiring

Partial snippet

Working example for Wiring.

import type { AudienceContext, GenerateObject } from "@absolutejs/audience";
import { meteredGenerateObjectAI } from "./usage/meteredAI";
import { aiProvider } from "./integrations/aiProvider";
import { embedTexts } from "./integrations/ragStore";

export const audienceCtx = (userSub?: string | null): AudienceContext => ({
  embed: (texts) => embedTexts(texts, "passage"),
  generateObject: ((req) =>
    meteredGenerateObjectAI({ ...req, provider: aiProvider, userSub })) as GenerateObject,
});

#Measuring audience overlap

Partial snippet

Working example for Measuring audience overlap.

import { profileAffinity, affinityOverlap } from "@absolutejs/audience";

const me = await profileAffinity({ name: "Me", signals: { niche, offer } }, audienceCtx());
const them = await profileAffinity({ name: company, signals: { summary, industry } }, audienceCtx());

const { score, sharedTopics, rationale } = affinityOverlap(me, them);
// score ∈ [0,1] — a *measured* overlap to use wherever you'd otherwise
// have an LLM guess (e.g. a Trust & Fit "audience overlap" dimension).

#Quick Start

Partial snippet

Build an AudienceContext from whatever you already use for structured generation and embeddings, extract two affinity profiles, then compare them without another model call.

import type { AudienceContext } from '@absolutejs/audience';
import { affinityOverlap, profileAffinity } from '@absolutejs/audience';

const ctx: AudienceContext = {
	embed: (texts) => embedTexts(texts, 'passage'),
	generateObject: (req) => generateObjectAI({ ...req, provider })
};

const me = await profileAffinity(
	{ name: 'Me', signals: { niche, offer } },
	ctx
);
const them = await profileAffinity(
	{ name: company, signals: { industry, summary } },
	ctx
);

const { rationale, score, sharedTopics } = affinityOverlap(me, them);
// score ∈ [0,1] — a measured overlap, not an LLM guess.

#Source Adapters

Source adapters implement the AudienceSource contract against third-party audience-intelligence APIs (Apache-2.0), putting measured data behind the same interface as the derived, own-it primitives.

@absolutejs/audience-audiensev0.0.1Audiense Insights adapter — reports, demographics, influencer/brand affinities, and measured audience overlap from Audiense’s proprietary social graph. Every AudienceSource method is optional, so consumers feature-detect and fall back to affinityOverlap when a method is absent.

Alpha status

Very early 0.0.x release — the AudienceContext and affinity type surfaces may still change between versions.

Measured, not guessed

Use affinityOverlap wherever you would otherwise have an LLM guess an overlap — for example the audience-overlap dimension of a @absolutejs/partnership Trust & Fit score.

#API reference

Search the declarations exported by the current package type files. Expand a symbol to inspect its source-backed signature.

23 symbols

AIMessageexportPermalink

AIMessage

Exported from @absolutejs/audience

Use this API in an outcome:Build commerce growth

npm GitHub

Continue toward an outcome

These playbooks show where this package fits, how to verify the combined system, and what changes before production.

Build commerce growthMove from discovery and enrichment through audience selection, consent-aware outreach, checkout, and reconciled money movement.

Current package surface

What ships today

@absolutejs/audiencev0.0.4 · betaCommerce & Growthnpm Source

3entry points24symbols

Import surface · click to copy

PreviousEnrich NextPartnership