The Mission: Cut through the AI hype and prove what actually works.
The Plot
The AI landscape is noisy. Everyone is talking about "Agents," but very few are actually shipping them. Prosus needed to stake their claim not just as investors, but as technical thought leaders. They didn't want a generic webinar; they wanted a movement.
The Hook: "The Office" Meets MLOps
We didn’t launch with a press release. We launched with a vibe.
To generate buzz, we dropped a "The Office" style mockumentary featuring MLOps founder Demetrios Brinkmann. The video followed Demetrios wandering the Prosus offices, blending painful second-hand embarrassment with eccentric character dynamics.
It wasn’t just funny; it was a strategy. By stripping away the corporate veneer, we signaled to the community that this wasn’t another stiff enterprise conference. It was a place for real people solving real problems. The video went viral in our niche, driving traffic, hype, and a flood of registrations.
The Main Event: A Two-Part "Mullet" Strategy
We designed a "Mullet" event architecture: Digital scale in the front, physical connection in the back.
Part 1: The Virtual Blitz (Nov 18)
16,951 Registrants | 4,124 Live Attendees | 37 Speakers
We turned a standard livestream into a variety show. To prevent "Zoom fatigue," we kept the energy chaotic but controlled:
- The Content: Hard-hitting technical talks from industry giants like OpenAI, Google, Meta, and Sierra. We tackled the unsexy but critical topics: MCP security, orchestration layers, and agentic commerce at scale.
- The Vibe: No stiff suits. Our host Demetrios anchored the stream with costume changes—morphing from a rockstar to a gorilla eating a banana, to a surfer—and playing guitar recaps of the talks live.
- The Result: Engineers stayed for the technical depth (Chip Huyen’s keynote on "Coding with AI" was a standout) but kept watching for the entertainment.
Part 2: The In-Person Deep Dive (Nov 19)
AI House, Amsterdam | 347 Curated Attendees
If Day 1 was about reach, Day 2 was about depth. We took over AI House in Amsterdam for an exclusive, curated expo.
- Headline: Mert Öztekin (CTO, Just Eat) breaking down how Europe’s largest delivery platform uses agents.1
- Fireside: A candid, "no-BS" discussion on the reality of Agent deployment with leaders from Arcade.dev and Prosus.2
- Engagement: We converted the digital momentum into 25 physical partner booths, turning passive viewers into active leads, demos, and handshakes.
The Podcast Series
Format: Candid, "Fly on the Wall" Conversations Scope: 17 Episodes | 226,300+ Impressions
While the world was screaming about AI hype, we wanted to whisper the truth. We launched a dedicated podcast series separate from our events—a collection of intimate, unscripted conversations between MLOps founder Demetrios Brinkmann and the varied experts within the Prosus ecosystem.
The Vibe: Unpolished & Real We ditched the corporate talking points. These were designed to feel like you were eavesdropping on two engineers grabbing a coffee and debating the hard stuff.
The Focus: Niche "Pockets" of Innovation We didn't just talk "AI." We dissected specific pockets of the Agentic Workflow space:
- The Reality Checks: "The Challenge with Voice Agents" (6.7k+ views)
- The Deep Dives: Exploring "Web Agents," "SQL Data Analysts," and "Enterprise Adoption Challenges."
The Result: This built a steady drumbeat of credibility. It wasn't about mass reach; it was about the right reach—building trust with engineers who can smell marketing fluff from a mile away.
Podcast Performance Snapshot:
- 32,800+ Short-form views
- 226,300+ Impressions
- 35,200+ Video views
The Verdict: Total Market Impact
We didn't just host an event; we built a community ecosystem.
- We Proved It: We moved beyond hype to show exactly how companies like Loblaws and iFood are shipping AI.
- We Scaled It: 16,951 registrations is a massive signal of market interest.
- We Humanized It: By mixing high-level engineering with humor and storytelling, we created a brand affinity that traditional B2B marketing simply cannot replicate.
Total Impact (one view)
