Voxluma.app Reviews and Details

Questions & Answers

1. What makes Voxluma.app unique?

   Most AI marketing tools give you one general-purpose chatbot and leave the marketing strategy to you. Voxluma gives you four specialists — Content, Leads, SEO and Analytics — that each already know your industry, location, tone and goals, so you get usable output instead of a blank prompt box. Paid plans add Clara, a document assistant that reads your contracts and leases in plain English. All of it for $9.99/month, priced for a business that doesn't have a marketing budget.

2. Why should a person choose Voxluma.app over its competitors?

   Tools like Jasper, Copy.ai and Hootsuite are built for marketing teams, and their pricing and complexity reflect that. Voxluma is built for the owner who is the marketing team. There's no seat-based pricing, no campaign-builder to learn, and no setup project — you tell it about your business once and the agents work from that. It also covers the whole loop, not just writing: finding leads, researching local SEO and reading your campaign numbers back to you in plain English.

3. How would you describe the primary audience of Voxluma.app?

   Small business owners and solo operators in the US — local services, trades, salons, childcare, fitness, restaurants and one-person consultancies. People running the business and the marketing at the same time, with no agency, no marketing hire, and no time to learn a marketing platform.

4. What's the story behind Voxluma.app?

   Voxluma was built by a small business owner who kept running into the same wall: the marketing hour never arrives. Agencies cost more than the marketing is worth at that scale, and the AI tools on the market were priced and designed for teams. So Voxluma was built the other way around — start from the owner doing everything alone, make the agents know the business so the output is usable straight away, and price it so it's an easy yes. The name is "voice in the light": your business voice, finally visible.

5. Which are the primary technologies used for building Voxluma.app?

   Front end — hand-written vanilla JavaScript (ES2020+), semantic HTML5 and hand-written CSS. No React, Vue or Svelte; no Tailwind or Bootstrap; no bundler or transpiler. The app ships as a single index.html with inline script and style, using CSS custom properties for the design system and CSS grid/flexbox for layout, with Cormorant and DM Sans for typography. The result is a first paint with zero framework payload.

   Back end — Cloudflare Pages for hosting and Cloudflare Pages Functions for the API layer: serverless JavaScript on the Workers runtime, executing at the edge close to the user, with no server to provision or patch. The Functions hold every secret — API keys never reach the browser — and each agent request is authenticated, metered against the user's plan and rate-limited server-side before it goes anywhere.

   Data and auth — Supabase: Postgres with row-level security for per-user data isolation, plus Supabase Auth for email and Google sign-in. The browser talks to Supabase through the official supabase-js v2 client under RLS; privileged writes happen only inside Pages Functions using a service-role key.

   AI layer — the agents run on Anthropic's Messages API, but the substance sits in the layer we wrote on top of it. Each of the four agents has its own purpose-built skill: a self-contained body of domain expertise describing how a specialist in that discipline actually works — what to establish before producing anything, which frameworks to apply, how the output should be structured, and the mistakes to avoid. The Content skill covers platform-native formats and hook construction; the Leads skill covers qualification criteria and outreach sequencing; the SEO skill covers local search intent, keyword clustering and brief building; the Analytics skill covers which metrics actually matter at small-business scale and how to turn them into a next action. Clara carries her own skill for document comprehension and risk flagging. At request time each skill is combined with the user's stored business profile — industry, location, tone, goals — so every response carries both discipline expertise and business context. Industry guardrails live in the same layer and are validated server-side.

   Document handling — Clara parses uploads entirely in the browser using self-hosted builds of pdf.js (PDF) and mammoth.js (.docx). Only extracted text is sent to a metered Function, nothing is persisted, and there are no document tables in the database by design.

   Analytics — Cloudflare Web Analytics: cookieless, privacy-preserving, no third-party tracking scripts.