AQuA · Launch Video Breakdown: Hook, Pacing & Motion Design
Recursive self-improvement for quantitative research.
Scene-by-scene timeline & spoken transcript
The Hook
Introducing AQuA: The Self-Improving Agent
“Meet Aqua, a recursively self-improving agent that works like a quantitative researcher, discovering factors, training models, and choosing what to test next.”
- On screen
- SYNTHETIC RESEARCH DEMO - NOT INVESTMENT ADVICE AQuA RECURSIVELY SELF-IMPROVING DISCOVER FACTORS TRAIN MODELS CHOOSE WHAT TO TEST NEXT
- Camera
- Static, centered framing on the Aqua character and expanding UI elements.
- Motion
- Animated UI elements appear and connect to the central character with subtle glows and line animations. Text fades in and out.
Product Reveal
Market Ideas Become Computable Factors
“It turns market ideas into factors, tests them across history, and trains models to decide which signals to trust.”
- On screen
- SYNTHETIC RESEARCH DEMO - NOT INVESTMENT ADVICE A MARKET IDEA BECOMES COMPUTABLE STOCK A STOCK B SAME RETURN • DIFFERENT PATH MEASURE CONSISTENCY factor = gain × consistency ONE RULE EVERY STOCK • EVERY DATE TESTED ACROSS HISTORY THE MODEL LEARNS WHEN TO TRUST EACH SIGNAL CALM MARKET VOLATILE MARKET TREND REVERSAL LEARNED MODEL PREDICTION VOLUME VOLATILE: rebalance trust TRUST CHANGES WITH CONTEXT
- Camera
- Static, centered framing on evolving data visualizations and flowcharts.
- Motion
- Data visualizations (candlestick charts) animate in. Flowchart elements appear sequentially with connecting lines and text highlights. The Aqua character subtly animates.
Feature Teaser
Recursive Learning and Architecture Search
“Every formula, score, and lesson stays in memory. Each result becomes the next question: new factors, longer windows, new model structures. Aqua iterates over different model architectures and finds the right one from the data.”
- On screen
- SYNTHETIC RESEARCH DEMO - NOT INVESTMENT ADVICE NOTHING IS THROWN AWAY factor = gain × consistency research score 0.52 trust changes with context MEMORY BELIEFS POLICY FORMULAS • SCORES • LESSONS EACH RESULT BECOMES THE NEXT QUESTION PLAN GENERATE UPDATE EVALUATE TRAIN + TEST RESULT feeds the next plan NEW FACTOR LONGER WINDOW NEW MODEL ARCHITECTURE SEARCH same market sequence • different structures LSTM MARKET SEQUENCE memory PREDICTION RECURRENT MEMORY MAMBA SELECTIVE STATE SCAN TRANSFORMER TRAIN → EVALUATE SEARCHING MODEL STRUCTURES
- Camera
- Static, centered framing on abstract data flow diagrams and model architecture comparisons.
- Motion
- Flowchart elements animate in sequence, highlighting connections and data flow. Text appears with subtle fades and scale animations. Model architecture names (LSTM, MAMBA, TRANSFORMER) animate in.
Call to Action
AQuA: Continuous Research Improvement
“The search stays open. The evidence stays fixed. And Aqua learns how to research better.”
- On screen
- SYNTHETIC RESEARCH DEMO - NOT INVESTMENT ADVICE FIXED EVIDENCE OPEN SEARCH NEW FACTOR LONGER WINDOW NEW MODEL AQuA LEARNS HOW TO RESEARCH BETTER
- Camera
- Static, centered framing on a conceptual diagram with the Aqua character.
- Motion
- Lines and nodes animate to form a conceptual diagram. The Aqua character appears with a subtle bounce. Text animates in with a clean fade and scale.
Motion design primitives
Synthesizable in Remotion · key components of the recursive-agent-explainer template.
AnimatedFlowchart
Rectangular nodes and connecting lines appear sequentially, often with a subtle glow or color change to indicate activation or data flow. Text within nodes fades or scales in.
+ Code construction recipe
Use `spring()` for node and line opacity/scale animations. SVG path animations for drawing lines. `interpolate()` for color changes on activation. React state to manage sequential appearance.
CharacterUIIntegration
A central character (Aqua) acts as an anchor, with UI elements (text, buttons, lines) emanating from or interacting with it. The character itself has subtle idle animations (e.g., blinking, slight bounce).
+ Code construction recipe
Render the character as an SVG or Lottie animation. Position UI elements relative to the character's coordinates. Use `spring()` for UI element entry animations and `loop()` for character idle animations.
DynamicTextHighlight
Key terms within sentences are highlighted with a color change or a subtle underline animation as they are spoken. On-screen titles and labels animate in with a clean, quick fade and slight scale.
+ Code construction recipe
Use `interpolate()` for text color changes based on audio cues or timeline position. `spring()` for title/label entry animations (opacity and scale). Implement a custom component for word-level highlighting.
Style DNA & aesthetic system
Visual aesthetic
Typography
Motion
Special effects
Pacing & rhythm
Conceptual story rhythm
- 1
Introduction to Aqua
Purpose. Introduce the core product (Aqua) and its primary function as a self-improving quantitative researcher.
Execution. The Aqua character is introduced centrally, with key functionalities (discover factors, train models, choose what to test next) animating around it.
- 2
Mechanism of Action
Purpose. Explain how Aqua processes market ideas into computable factors and trains models to trust signals.
Execution. Flowcharts and data visualizations (candlestick charts, signal processing diagrams) illustrate the conversion of market ideas into factors and the model's learning process.
- 3
Recursive Improvement
Purpose. Detail Aqua's continuous learning loop, memory retention, and ability to search for optimal model architectures.
Execution. Diagrams showing memory storage, a cyclical feedback loop (plan, generate, train, evaluate, update), and a comparison of different model architectures (LSTM, MAMBA, Transformer).
- 4
The Core Advantage
Purpose. Summarize Aqua's unique ability to continuously improve its research methodology.
Execution. A final conceptual diagram emphasizing 'Open Search' and 'Fixed Evidence' with the Aqua character, reinforcing its learning capability.
Remix in Animatiq Studio
Master remix prompt · agent-ready
Sample remix
Objective
Explain 'Synapse AI', an AI agent that automates complex legal document analysis and contract generation.How the recipe adapts
The 'AnimatedFlowchart' primitive would illustrate Synapse AI's workflow: 'Ingest Documents' -> 'Identify Clauses' -> 'Extract Entities' -> 'Generate Draft Contract' -> 'Review & Refine'. The 'CharacterUIIntegration' primitive would feature a new AI character, 'Lexi', interacting with these flowchart elements, perhaps with thought bubbles showing legal concepts. 'DynamicTextHighlight' would emphasize terms like 'clauses', 'entities', and 'contract generation' as they are narrated. The color palette would shift to deep blues and purples with bright yellow-green accents, maintaining the tech-forward neon aesthetic. The overall pacing would remain slow and explanatory, allowing the viewer to absorb the complex legal process.Remix blueprint · recursive-agent-explainer
Scene durations (5 scenes, 48s)
Key Remotion components
Related AI Agents Product Launches
Explore all AI Agents launches →
To ensure that artificial general intelligence benefits all of humanity
Gopuff introduces Go, an AI shopping assistant built with SpaceXAI: say what you need and the order is placed.
Iliad (Troy) trailer made by Grok Imagine 1.5, which was just released