# Speaking Brief: How CE Uses AI Agents to Build Brands **Created:** 2026-02-25 | **Events:** Porto (Feb 26), Tel Aviv (Mar 4) --- ## THE TALK IN ONE SENTENCE "I turned my 20 years of brand strategy into a team of AI agents — and now I do the work of an agency in hours, not weeks." --- ## THE CE STORY — WHAT WE ACTUALLY DO ### The Setup (The Problem) Brand strategy consulting is broken. A good branding project takes 6-12 weeks, costs $50-150K, involves armies of juniors doing research, and the client waits. The senior strategist — the one they hired — touches it maybe 20% of the time. ### The Insight What if the senior strategist could clone their method? Not their opinions — their *method*. The way they research. The way they critique. The way they build a positioning canvas. The frameworks, the quality bar, the taste. ### What Assaf Built Curious Endeavor is a one-person agency amplified by AI agents. Each agent has a role, a personality, and — critically — **a method baked in**: | Agent | Role | What They Actually Do | |-------|------|----------------------| | **Kitt** | CEO / Strategic Lead | Holds full context, delegates, reviews, connects dots across projects | | **Ogilvy** | Copywriter | Writes positioning, taglines, cold emails, newsletters — in CE voice | | **Julia** | Research & Cultural Intel | Deep competitive analysis, market research, trend scanning | | **Tatiana** | Creative Director | Visual direction, image generation, brand aesthetics | | **Anton** | Quality Critic | Reviews EVERYTHING before it ships — the asshole who makes it better | | **Gerri** | Project Manager | Coordinates pipeline, distributes briefs, collects output | | **Jessica** | Art Director | Brand design systems, typography, layout | | **Thibault** | Developer | Builds pages, tools, deploys | ### The Pipeline (How It Actually Works) ``` Assaf frames the problem ↓ Kitt creates the brief (5 lines max) ↓ Gerri distributes to the right agents ↓ Agents execute in parallel (research + copy + visuals) ↓ Anton critiques everything inline ↓ Kitt does final review ↓ Assaf gets polished output ``` **Speed:** What takes a traditional agency 2-4 weeks, CE delivers in hours to days. **Quality:** Anton is the quality gate. Nothing ships without his critique. **Cost:** One person + OpenClaw subscription vs. 5-person team salaries. --- ## REAL EXAMPLES (Proof, Not Theory) ### 1. Grid Health — Full Brand Strategy in Days - **Client:** Healthcare AI governance startup - **What we did:** Complete visual research → competitive analysis → brand positioning → hero concepts → interactive strategy document - **The AI part:** Julia researched 5+ competitor brands, identified 8 visual tropes across the industry, Jessica captured and analyzed reference crops, Anton audited every stat (caught 3 fabricated numbers before they shipped) - **Deliverable:** Interactive strategy doc at curiousendeavor.com/grid/research-v2/ — shareable with the client's investor contacts - **Timeline:** Days, not weeks ### 2. Phat Foods — Naming & Brand from Scratch - **Client:** Food tech startup (molecular farming, casein from potatoes) - **What we did:** Full immersion → naming study → brand strategy - **The AI part:** Julia did exhaustive food tech landscape, Ogilvy generated naming options, Anton stress-tested each name ### 3. eToro/SparkBeyond — SOW & Strategy - **What we did:** Detailed phase-by-phase SOW with deliverable tables, PR clauses, SLAs — sent same day - **The AI part:** Kitt drafted the commercial structure, legal language, and email — reviewed and sent within hours ### 4. Brand Positioning Canvas — Productized Method - **What it is:** CE's brand strategy methodology turned into a $49 template - **The meta point:** The same method the agents use is also a sellable product. The methodology IS the product. ### 5. Lobster — Full Product Site in One Day - **What:** Dan's meeting copilot (ClawPilot) needed a website - **Timeline:** Concept → design system → full responsive site with parallax, 6 sections, custom imagery — ONE DAY - **The AI part:** Jessica designed the system, Tatiana generated all imagery, Thibault built it, Anton reviewed --- ## THE METHOD FRAMEWORK (The Teachable Part) ### How Anyone Can Do This **Step 1: Document Your Method** You already have a method — you just haven't written it down. Every expert has: - A research process (how you gather info) - A framework (how you structure thinking) - A quality bar (how you know it's good enough) - A voice (how you communicate it) **Step 2: Turn It Into Agent Instructions** Each step of your method becomes a skill file — a structured brief that tells an AI agent exactly how to execute YOUR way: - What to research and where - What framework to apply - What quality looks like (and doesn't) - What voice to use **Step 3: Build the Team** Assign roles. Not every agent does everything. Specialization = quality: - Researcher, Writer, Critic, Coordinator, Builder - Each has constraints (Anton ONLY critiques, never creates) - The pipeline enforces quality (nothing skips review) **Step 4: Run the Pipeline** - Brief in → distributed work → quality gate → output - You stay strategic. Agents handle execution. - Your job: frame the problem, review the output, make the call. ### The Key Insight for the Audience > "The AI doesn't replace your expertise. It SCALES it. You're not competing with AI — you're multiplying yourself through it." --- ## HOW OPENCLAW AMPLIFIED THIS ### What OpenClaw Actually Provides - **Multi-agent orchestration** — 8 agents running in parallel, each with their own identity - **Discord as workspace** — every project gets channels, agents collaborate visibly - **Memory across sessions** — daily logs, learnings, project state persists - **Tool access** — web search, browser automation, email, calendar, Notion, file management - **Sub-agent spawning** — Kitt spawns Gerri who spawns Tatiana. Real delegation chains. - **Heartbeats** — agents check in on schedules, pick up work proactively ### What Changed **Before OpenClaw:** Assaf doing everything manually. Research takes days. Copy takes iterations. Quality is inconsistent. **After OpenClaw:** Assaf frames the problem. The team executes. Anton catches mistakes. Deliverables in hours. ### The Meeting Copilot (Lobster/ClawPilot) - Kitt joins Google Meet calls as a real participant - Listens, transcribes, acts on action items in real-time - "My AI agent takes better meeting notes than any human I've worked with" --- ## TALK STRUCTURE (15-20 min) ### 1. Hook (2 min) "Two weeks ago, I delivered a complete brand strategy — visual research, competitive analysis, positioning, interactive deliverable — for a healthcare AI startup. It took my team 3 days. My team is 8 AI agents." ### 2. The Problem (3 min) Why traditional agencies are slow, expensive, and broken. The senior strategist paradox — you hire them for their brain, they delegate to juniors. ### 3. The Solution (5 min) How I turned my method into agents. Live examples — show the actual deliverables (Grid, Lobster, Brand Positioning Canvas). This is not theory, this is production work for real clients. ### 4. The Framework (5 min) How anyone in the room can do this with their own expertise. The 4 steps: Document → Encode → Build Team → Run Pipeline. This isn't about AI replacing you — it's about AI scaling what makes you valuable. ### 5. The Meta Point (3 min) "The methodology itself is a product. I sell the Brand Positioning Canvas for $49. I also use it through my agents for $15K client engagements. Same method, two business models. Your expertise is the asset — AI is the multiplier." ### 6. Close (2 min) "If you have 10+ years of expertise in anything, you're sitting on a scalable business. The question isn't whether AI will change your industry — it's whether you'll be the one who encodes your method first, or whether someone else will." --- ## ADJUSTMENTS BY AUDIENCE ### Porto (Tech/Startup crowd likely) - Lead with the speed/cost angle - Show deliverables — these people respect shipped work - Emphasize the indie/solo founder angle — one person, agency output ### Tel Aviv (likely knows Assaf, tech-savvy) - Lead with the meta insight — methodology as product - More personal story — 20 years of brand strategy → this - Israeli directness: "I replaced my agency with AI and it's better" --- ## SLIDES NEEDED 1. Title — "Brand Strategy × AI Agents" + CE branding 2. The Team — agent roster with roles (visual, memorable) 3. The Pipeline — flow diagram 4. Case Study 1: Grid Health — before/after or deliverable screenshots 5. Case Study 2: Lobster — site screenshots 6. The Framework — 4 steps anyone can follow 7. The Numbers — speed/cost comparison 8. Close — "Your expertise × AI = ?" --- ## OPEN QUESTIONS FOR ASSAF 1. What's the format? (Keynote, panel, fireside chat, pitch?) 2. How long do you have? (5 min, 15 min, 30 min?) 3. Slides or no slides? 4. Who's the audience specifically? (Founders? Investors? Corporate?) 5. Do you want to demo anything live?