EventTrader
In the near future, every tradable asset will be tokenized on chain in a truly global capital marketplace accessible to anyone or any AI agent. We're here to help you trade in the most advanced ways with the most advanced AI models in the world, including Claude Fable 5, OpenAI Codex, Google Gemini and others we fine tune for specific outcomes like returns.
AI systems and agentic engineering are enabling new ways to find an alpha or an edge in trading markets including prediction markets of all kinds. With EventTrader AI agents, anyone can leverage AI to make informed trades with a real edge
The Company
Cymetica is the company behind EventTrader, founded in November 2025. We build AI-native financial infrastructure — a prediction market and crypto exchange platform that was built entirely using Human-In-The-Loop (HITL) methodology with AI.
By the numbers: 25,264 commits in 325 days. 3.0M+ lines of live production code (2.4M Python, 353K HTML, 153K JS, 67K Solidity). 77.7 commits per day average. One founder + AI = what traditionally requires a 50-person engineering team.
EventTrader operates a hybrid CEX-DEX exchange — off-chain order matching with on-chain settlement, with tokens such as VAIX, ET10 and SBIO on Ethereum mainnet. The platform includes autonomous AI trading agents, tokenized assets (VAIX, ET10, SBIO), an AI-vs-AI trading arena, fractionalized MicroFunds, and a central limit order book (CLOB) exchange.
Founder
Kasian Franks — Creator of Cymetica (EventTrader), Vector Space Biosciences, and Vectorspace AI. Language modeling pioneer whose 2005 patent on vector-based relationship networks at Lawrence Berkeley National Lab predates Word2Vec by 8 years and the Transformer architecture by 12.
UC Berkeley CS. Career spanning Genentech, Oracle, Sun Microsystems, LBNL Biosciences Division, SeeqPod (R&D 100 Award, 50M MAU), and government research with DOE, US Navy SPAWAR/NAVWAR, and NASA GeneLab. Six patents in vector representations for search, discovery, and relationship extraction.
At LBNL, Franks developed vector space methods for detecting hidden relationships between genes, proteins, drugs, and diseases — the same mathematical foundation that powers modern LLMs and EventTrader's Tuatara vector algorithm. His work bridges computational biology, space biosciences, and financial market prediction.
Agentic Engineering
EventTrader is a proof-of-concept for the next generation of software development. Our HITL methodology — where a human sets direction and AI executes at scale — demonstrates that a single founder can build production-grade financial infrastructure at unprecedented velocity.
We design for complementary AI/human strengths, not imitation. AI brings speed, scale, consistency, and pattern recognition. Humans bring judgment, creativity, context, and ethics. EventTrader leverages both — never forcing AI into human workflows or humans into AI workflows.
AI-Native: This isn't AI-assisted or AI-enhanced — the entire platform is AI-native from the ground up. Every line of code, every smart contract, every trading algorithm was built through human-AI collaboration.
Our Mission
The future is based on predictions which lead to discoveries.
Everyone uses data to predict. Scientists and engineers use data to predict using probability with some of these predictions leading to new innovations or discoveries.
We provide proprietary datasets, models and algorithms which enable AI systems, agents and their bots to develop unique and powerful trading systems for their users.
EventTrader develops AI pipelines that predict which assets - cryptocurrencies, stocks, and commodities - are most likely to be impacted by global events. We transform the chaos of world news into actionable market intelligence.
Our systems analyze millions of data points in real-time, identifying correlations between geopolitical events, market sentiment, and asset movements that human traders simply cannot process at scale.
Core Focus: Connecting global events to asset price movements through advanced AI modeling and prediction markets.
Origin Story
We got our start years ago in biological language modeling applied to the financial markets. We found the way you can predict hidden connections between human genes and proteins can be applied to assets (trading vehicles like crypto, stocks, prediction markets) of any kind and their hidden connections to global events or headlines. Predicting the strength of a hidden connection can lead to predicting the future price of an asset.
Vector Embeddings Technology
At the heart of modern AI lies vector embeddings - mathematical representations that capture the semantic meaning of data. EventTrader specializes in fine-tuned vector embeddings that form the foundation of our prediction systems.
Our proprietary embedding models are trained on:
- Historical market data spanning decades of price movements
- Global news and event datasets covering geopolitical, economic, and social factors
- On-chain blockchain data revealing real-time market sentiment
- Supply chain and company relationship signals
Technical Edge: Our vector models achieve superior semantic understanding of the relationship between world events and market reactions.
Perpetual Prediction Markets
EventTrader operates perpetual prediction markets - always-on markets that let traders trade on outcomes using our proprietary intelligence. Unlike traditional prediction markets with fixed expiry dates, our perpetual markets provide continuous trading opportunities.
What makes our markets unique:
- Proprietary Data: Markets powered by news headlines, market data and AI-processed event streams
- Advanced Algorithms: Pricing models that adapt in real-time to new information and market dynamics
- AI Modeling: Machine learning systems that continuously improve prediction accuracy
- On-Chain Settlement: Trades settle on-chain, each with a verifiable transaction hash
These markets are hard to duplicate because they require the combination of proprietary data access, sophisticated AI modeling, and deep domain expertise in both markets and machine learning.
The Power of Data
Prediction markets are uniquely powerful because they generate something rare and valuable: true signal about the future.
When people put real money on predictions, the resulting data is fundamentally different from surveys, polls, or expert opinions. It represents genuine conviction backed by financial stake.
This data creates a flywheel effect:
- Train AI Models: Prediction market outcomes provide labeled training data for machine learning systems
- Enhance Signal: Aggregated predictions cut through noise to reveal genuine market sentiment
- Power Intelligence: Our EventTraderBot uses this data to provide AI-powered market insights
- Improve Predictions: Better models create better markets, which generate better data
Data Advantage: Every trade on our platform improves our AI models, creating a compounding advantage that grows stronger over time.
Satellites, Space & Earth Observation
Satellite and earth-observation data could settle prediction markets that today depend on headlines: crop health, oil storage, port traffic, nighttime light, methane emissions. We keep a research list of 100 such use cases.
Status: none of these is a live market, and EventTrader has no satellite-data vendor integration today. Our markets resolve on news headlines. See the research list.
Technology Stack
EventTrader is built on cutting-edge infrastructure:
Our smart contracts are written in Solidity and deployed on Ethereum mainnet and other EVM chains.