Hey HN,

I'm Zack, CEO at Averi. We just launched what we're calling The AI Marketing Workspace—a complete rebuild of how we think marketing teams should work with AI.

Most AI marketing tools solve for one narrow use case:

-AI copywriting tools that spit out generic content -Strategy platforms that stop at recommendations -Analytics dashboards that show what happened but don't help you act

Our Approach: Workspace > Tool

Instead of building another point solution, we built a workspace that enables complete marketing workflows: Plan → Create → Execute → Scale.

What that means:

Plan: AI-powered strategy and planning with your brand context built in. Activate expert strategists when needed. Create: /create mode takes briefs to deployment-ready assets. Collaborate with specialists in real-time. Execute: Deploy with full context preserved. Experts see your strategy, drafts, and brand. Scale: Library captures what worked. Every project trains the AI and makes future work faster.

What We Built

/create Mode: Three-Phase Creation Engine The execution system at the heart of the workspace.

Discuss Phase: Brief the AI with contextual questions specific to each asset type

Draft Phase: AI generates first draft using Synapse architecture (routing between AGM-2, GPT-5, Claude based on task) One-click adjustments to length, vibe, or language Full draft regeneration with preserved context

Edit Phase: Standard text editor with AI power-ups Highlight any text → regenerate with AI Add AI-generated paragraphs, images, infographics between sections Full collaborative editing

Output: Save as .AVRI files (native format maintaining full editability and AI context) Download in standard formats Share with experts in the workspace for refinement

The Workspace Architecture

Brand Core: Central repository for mission, values, brand voice, ideal customers, goals. Trains the AI and provides persistent context across all interactions.

Library System:

System Folders: Marketing Plans, ICPs, Products, Saved AI Conversations Custom Folders: Organize content however you want, train AI on specific contexts .AVRI files open directly in edit phase for seamless collaboration Expert Network Integration: Type /intro to request vetted marketing specialists. Review profiles, activate experts, collaborate in the workspace.

What Makes This Different

1. Complete workflow, not isolated tasks: Most tools solve for one phase (strategy OR creation OR execution). Averi enables the full flow with persistent context throughout. 2. Context-aware by design: The AI doesn't just respond to your prompt. It knows your brand, your active plans, your audience, your goals—and maintains that context. 3. Execution-first: We don't stop at suggestions or strategy. The workspace takes you from first thought to deployment-ready assets to measurable results. 4. AI + Human, built in: Expert collaboration isn't a separate marketplace—it's native to the workspace. 5. Compounding institutional memory: Library and .AVRI files create a feedback loop. Every successful project makes future work better.

Open Questions We're Exploring

1. Workspace vs. Tool Mental Model: How do you get users to think of this as "where they work" vs. "a tool they use for 10 minutes"? We've found the multi-tab interface helps, but curious about other approaches. 2. File Format Ecosystem: We're betting that .AVRI files (maintaining full AI context + edit history) become a standard for marketing collaboration. 3. Self-Serve Conversion: We just opened self-serve today. Historically we were demo-driven. Interested in what others have learned about freemium → paid. 4. Integration Strategy: We're building integrations to CMS, email platforms, social schedulers. Question: do users want to push directly from Averi, or prefer to download and upload themselves?

Try It

Thanks! Zack

Ask HN: Why does HN keep "deleting" my posts?

A Other Word Game, Inspired by the New Yorker's Shuffalo

Quantum Foam SE

AI Uses Functions to Fetch Real Data (Not Just Chat)

Norway's Public Buses Have a Chinese Backdoor No One Knew About

StackRender: From an Idea to production-ready database in no time

Show HN: Gaggle – A DuckDB extension for working with Kaggle datasets

Accessibility score is lying to you

Neural implant smaller than salt grain wirelessly tracks brain

Show HN: I made a lightweight way to send tasks to others

Our response to Disney's request to restore ABC

The Art of Atari (2016)

Sacrificing the Present for an Anticipated Memory

The Paranoid Guide to Running Copilot CLI in a Secure Docker Sandbox

AI Will Flatten Workforce Inequality–If We're Honest About What That Means

Agents 2.0: From Shallow Loops to Deep Agents

Building Coding Agent in Rust – Implement Chat CLI – Part-2 [video]

I Will Not Be Enabling Full Null Support in Adobe ColdFusion 2025

Apple App Store

Thanks for Subscribing

Show HN: Averi – The AI Marketing Workspace

Roger Penrose, the Outlier Who Keeps Asking "What's Missing in Quantum Physics?"

Ranking LLMs based on 180k French votes (French government's AI arena)

Hosting an ATProto PDS without containerization

Dick Cheney, Powerful Vice President and Washington Insider, Dies at 84

RV There yet Guide

A Challenge to Roboticists: My Humanoid Olympics

Roger Penrose: Why the Big Bang Was Not the Beginning [video]

I Processed the Internet on a Single Machine to Find Valuable Expired Domains

Microservices? No, modularity is what matters

Show HN: AI Invoicing with Jinna – Create, Send and Chase to Get Paid Faster