Brand Voice Analyzer ingests your existing content — website copy, blog posts, emails, social posts — and extracts a detailed brand voice profile. It identifies your tone patterns, vocabulary preferences, sentence structure tendencies, and emotional register. The output is a reusable style guide that other agents (Content Machine, Ad Optimizer) can reference to match your voice exactly.
Capabilities
Analyzes tone across 12 dimensions: formality, humor, urgency, authority, warmth, etc.
Extracts vocabulary fingerprint — words you overuse, words you avoid, brand-specific terms
Maps sentence structure patterns: average length, question frequency, CTA placement
Generates a consistency score comparing your content across channels
Outputs a machine-readable style guide that other agents consume automatically
Identifies voice drift — where your recent content deviates from your core brand
Example Workflow
You say
“Analyze the brand voice at targetbrand.com — they are in the B2B SaaS space”
Agent does
Scrapes website copy, analyzes tone/vocabulary/sentence patterns, and compares against industry benchmarks
You get
A brand voice profile with tone scores, vocabulary list, style guide rules, and a consistency rating
Models Powering This Agent
Nemotron Ultra 253B (deep linguistic analysis)
Gemma 4 31B (pattern extraction + scoring)
All models run on NVIDIA NIM infrastructure. Model selection is automatic based on task requirements.
Safety & Verification
Every execution of Brand Voice Analyzer passes through a 5-layer safety pipeline before results reach you. No exceptions.
Jailbreak DetectionBlocks prompt injection before execution