ObsessionDB · Launch Video Breakdown: Hook, Pacing & Motion Design
The realtime data warehouse for agents, ClickHouse with compute and storage separated
Scene-by-scene timeline & spoken transcript
The Hook
Agent Data Appetite
“Your agents are hungry for data. Can your data warehouse feed that appetite?”
- On screen
- OpenAI Codex model: gpt-5.2-codex medium directory: ~/activity > Inspect recent customer activity | ILLUSTRATIVE TELEMETRY GPU usage 71% CPU usage 76% Temperature 55°C | event_type event_count page_view 9 report_open 9 export 9 query_run 9
- Camera
- Rapid sequence: terminal UI, chip macro, drone facility shot, database table
- Motion
- Fast UI cuts with telemetry HUD overlays
Problem Agitation
Exploding Queries & Infra Bills
“Millions of questions, billions of queries, running autonomously 24/7. As data compounds, so does your infra bill. And agents are just getting started.”
- On screen
- Millions of questions. | Billions of queries. | Running autonomously, 24/7. | Queries 016 SELECT count(*) FROM sessions; | Infra bill | Cursor Perplexity Codex Claude Grok Gemini
- Camera
- Dense text grids zooming out, scrolling telemetry logs, rising line chart against server rack
- Motion
- Kinetic typography grids, SVG path inflation chart, rotating logo constellation
Product Reveal
Introducing ObsessionDB
“That's why we built ObsessionDB, the fastest and most scalable data warehouse for AI. ObsessionDB is a hundred times faster while being 20 times cheaper than the existing data platforms like Snowflake or Databricks.”
- On screen
- ObsessionDB | 100x faster 20x cheaper vs. Snowflake Databricks
- Camera
- Close-up founder to medium three-shot founder panel
- Motion
- Direct camera delivery with clean on-screen benchmark typography
Feature Teaser
Decoupled Architecture & Scale
“We did this by taking the fastest analytical engine and applied compute storage separation to it. And with this, we achieved frontier performance. And it's live. We're already processing trillions of rows per day. We scaled our previous company to petabytes and know what it means to see your infra bill explode. That's why we set out to solve this problem. Your agents are just getting started. We're building ObsessionDB so your infrastructure can grow with them.”
- On screen
- One analytical engine. Compute and storage, separated. Compute Storage | Already in production Trillions of rows. Every day. | Infrastructure that grows with your agents. Agents Infrastructure
- Camera
- Intercut between founder explanations and minimal 2D vector architectural diagrams
- Motion
- Dynamic bar chart expansions, block cloning animations, typography highlights
Call to Action
Agent-Ready CTA
“Make your infrastructure agent-ready. Come talk to us at ObsessionDB.”
- On screen
- ObsessionDB obsessiondb.com
- Camera
- Founder close-up to trio shot, snapping to centered dark end-card
- Motion
- Logo build with typographic URL and link glyph
Motion design primitives
Synthesizable in Remotion · key components of the ai-infra-explainer template.
SeparationDiagram
Horizontal compute modules spawn and scale above a fixed storage foundation layer.
+ Code construction recipe
Render base storage bars in Remotion, then use spring-driven stagger to clone and distribute orange compute blocks horizontally.
InfraBillCurve
An orange line chart exponentially climbs against dark grid lines alongside scrolling SQL queries.
+ Code construction recipe
Animate SVG path strokeDashoffset mapped to interpolate(frame, [0, 90], [pathLength, 0], { extrapolateRight: 'clamp' }).ConstellationOrbit
Six AI company logos rotate along a central elliptical ring in continuous loop.
+ Code construction recipe
Map items along polar coordinates angle = (frame * speed + index * (360 / count)) * (Math.PI / 180).
Style DNA & aesthetic system
Visual aesthetic
Typography
Motion
Special effects
Pacing & rhythm
Conceptual story rhythm
- 1
The Appetite
Purpose. Frame AI agents as infinite data consumers.
Execution. High-velocity cuts through code, silicon, industrial telemetry, and queries.
- 2
The Cost Crisis
Purpose. Highlight the cost scaling bottleneck for AI teams.
Execution. Steep exponential infra curve climbing beside SQL logs.
- 3
The Solution & Credibility
Purpose. Present founders, architecture, and production proof points.
Execution. Trio founder interview intercut with compute-storage vector diagrams.
- 4
The Invitation
Purpose. Direct viewers to upgrade their infra.
Execution. Direct-to-camera address resolving to minimalist domain logo card.
Remix in Animatiq Studio
Master remix prompt · agent-ready
Sample remix
Objective
Launch promo for an open-source real-time vector search engine.How the recipe adapts
Swap SQL queries for embedding vector logs, show vector cluster diagrams expanding horizontally, and feature founders discussing latency reduction.Remix blueprint · ai-infra-explainer
Scene durations (5 scenes, 58s)
Key Remotion components
Related AI & ML Product Launches
Explore all AI & ML launches →
Introducing Griffin, the first AI to pass the video Turing test
The most emotive and controllable TTS model
We just released QVAC TranscribePsy, a speech-to-text model small enough to run inside a phone app. It covers 25 European languages on a laptop and 8 on a phone, and on an iPhone 17 Pro Max's Ne